THE FERN GAZETTE

VOLUME ELEVEN PARTS TWO AND THREE

1975

THE JOURNAL OF THE BRITISHPTERIDOLOGICAL SOCIETY FERN GAZETTE

VOLUME l1 PAR1S2 & 3 1975

CONTENTS

Page

A re-defintion of the Gymnogrammoid genus Austrogramme Fournier - E. Hennipman 61

The biogeography of endemism in the Cyatheaceae- R. Tryon and G Gastony 73 wnathyrium in the Azores- W.A. Sledge 81

The gametophyte of Chingia pseudoferox- Lennette R. Atkinson 87

Ta xonomic notes on some African species of Elaphoglossum - R.E. G Pichi Sermolli 95

Observations on the spread of the American fern Pityrogramma calomelanos -E.A. CL. E. ::Che/pe 101

A phytogeographic analysis of Choc6 Pteridophytes- D. B. Le/linger 105

Studies in the systematics of filmy ferns: I. A note on the identity of Microtrichomanes - K. Jwatsuki 115

A hybrid polypody from the New World tropics - W.H. Wagner and Florence Wagner 125

Aspidistes thomasii- a jurassic member of the Thelypteridaceae- }.£?. Lovis 137

A new arrangement for the pteridophyte herbarium - j.A. Crabbe, A.C jermy and }.M. Mickel 141

A note on the distribution of lsoetes in the Cadiz Province, Spain - Betty Moles worth A lien 163

Lecanopteris spinosa; a new ant-fern from Indonesia - A.C jermy and T. G Walker 165

Dryopteris tyrrhena nom. nov. - a misunderstood western Mediterranean

species - C.R. Fraser jenkins and T. Reichstein 177

THE BRITISH FERN GAZETTE Volume 11 Part 1 was published 6 February 1975

Published by THE BRITISH PTERIDOLOGICAL SOCIETY, c /o Department of Botany, Museum {Natural History} . London SW7 5BD. \

)�eA·r""-�\t". 197�· This issue is dedicated

to

RICHARD ERIC HOLTTUM

Honorary Member and Past- President

of the British Pteridological Society,

Director of the Singapore Botanic Gardens 1925- 1949

and

Professor of Botany,

Universit y of Malaya, Singapore 1949- 1954

on the occasion

of his

Eightieth Birthday

20th July 1975 FERN GAZ. 11(2 & 3) 1975. 61

A RE-DEFINITION OF THE GYMNOGRAMMOID GENUS AUSTROGRAMME FOURNIER

E. HENNIPMAN Rijksherbarium, Leiden, Netherlands

ABSTRACT The Gymnogrammoid genusAustrogramme is reinstated to accommodate five species formerly referred to Aspleniopsis, Rheopteris, and Syngramma. A key is included. The genus is divided into two sections including Section Aspleniopsis (Fournier) Hennipman. The following new combinations are proposed: Austrogramme asplenioides (Holttum) Hennipman, A. boerlageana (v.A.v.R.I Hennipman, A. decipiens (Metteniusl Hennipman and A. francii (Rosenstock) Hennipman. The now monotypic genus Rhaopteris Alston is excluded from the Gymnogrammoid ferns. The relationship of Austrogramme to Taenitis and Syngramma is discussed.

INTRODUCT ION When making routine identifications at the Rijksherbarium I came across a specimen of Rheopteris asplenioides Holttum (New Guinea) which appeared to me as probably intimately related to Aspleniopsis decipiens (Mettenius) Kuhn, which is indigenous to New Caledonia, the New Hebrides, and possibly also New Ireland (Brownlie 1969). This aroused my interest, also as the real affinity of the mo00typic genusAspleniopsis was said to be unknown (Copeland 1947). In a correspondence, Prof. Holttum suggested that Syngramma boerlageana v.A.v.R. from the Moluccas, and the New Caledonian Syngramma marginata (Mettenius) Diels and S. fra ncii Rosenstock, might be related to these as they did not fit into the genus Syngramma as presently construed by him. Detailed studies of the vascular organisation, the disposition of the sporangia, the spores, and the paraphyses, surprisingly showed conspicuous similarities in all of them. They further appeared to be different from Syngramma and Taenitis by the combination of leaf·shape and venation pattern. As a result, the genus Austrogramme Fournier, based on A. marginata is reinstated. The relationship of the Gymnogrammoid genus Austrogramme to Syngramma sensu Holttum (1954) (thus including Craspedodic tyum Copeland), and to Ta enitis sensu Holttum (1968) is discussed also in view of Walker (1968) and my own observations on representatives of the latter two genera. Another finding of interest regards the identity of the now monotypic genus Rheopteris Alston. Rheopteris cheesmaniae Alston - the type species of the genus­ appeared to be genetically different from R. asplenioides (here referred· to Austrogramme). The genus which was formerly referred to the Gymnogrammoid �erns by Holttum (1962) shows several aberr·ant features. Its relationship will be discussed in detail later (Hennipman, in preparation).

MATE RIAL AND METHODS The present study is based on herbarium material present at the Rijksherbarium (L), as well as on additional collections of Austrogramme asplenioides, A. decipiens, A. francii, A. marginata, and Rheopteris cheesmaniae present at the British Museum (Natu ral History), London (BM); the Royal Botanic Gardens, Kew (K), and the Museum National d'Histoire Naturelle, Paris (P). Spores were stu died embedded in glycerine gelatine with the light microscope, 62 FERN GAZETTE: VOLUME 11 PART 2 & 3 (1975)

and coated with gold in a Cambridge Stereoscan. The details of the spore-morphology are described from the electron micrographs. Paraphyses were studied after boiling in an aqueous solution of potassium hydroxide for one minute. Cross-sections of rhizome and petiole were made following routine procedures.

OBSERVAT IONS Vascular organisation The vascular organisation of Austrogramme is uniform. The rhizome is solenostelic like in Taenitis and Syngramma species. T�e petiole of Austrogramme sh ows on cross-section two lateral vascular bundles. In Taenitis the number of vascular bundles in the petiole was found to be one (V-shaped) in T. cordata (Van Royen 5462) and two in the other species. In Syngramma the number of vascular bundles is four inS. luzonica, three inS. alismifolia, and two inS. borneensis. The observations on Syngramma and Taenitis are generally in accordance with Walker (1968).

Shape of the leaves and venation pattern In Austrogramme compound leaves have a triangular terminal segment (A. ·asplenioides, A. boerlageana, A. decipiens), and a venation pattern which is either reticulate (A . boerlageana) or (largely) free (A. asplenioides, A. decipiens). The species with simple leaves (A. francii, A. marginata) have free veins. Compound leaves of Ta enitis and Syngramma have the terminal segment ±. conform to the pinnae. The venation pattern of all the species, thus including those with simple leaves, shows anastomosing veins at least towards the margin of the pinnae or the leaves. For illustrations of the venation patterns of Syngramma and Taenitis species see Holttum (1954) and Walker (1968).

The disposition of the sporangia In Austrogramme the sporangia are situated in .± orbicular to elongate sori, varying in size and generally situated, either mainly or exclusively along the distal parts of each vein or irregularly. lh Syngramma the sporangia are generally inserted regularly all along each vein. Sporangia are variously situated in Taenitis (for details see Holttum 1968), a condition as found in Austrogramme is, however, not realised.

Paraphyses In Austrogramme the paraphyses of the sori are mainly inserted either on the stalk of the sporangium (in groups of 2-8) or on the epidermis. They are uniseriate, unbranched, and vary in shape but usually have the upper part curved. They consist of a± conical, small, hyaline or brownish, glandular (?), terminal cell, generally 4-7, ± spherical or elongate, yellowish to brownish, glandular subterminal cells, along with a varying number of non-glandular ± hyaline cells. The number of non-glandular cells is generally 1-3 when the paraphysis is inserted on the stalk of the sporangium. In mature glandular cells of the paraphyses of A. asplenioides, A. boerlageana, and A. marginata, small openings were observed (figs 1-5). In Tae nitis the paraphyses show a range of variation. 'I n T. blechnoides, T. in te"upta, and T. marginata, they are mainly inserted on the stalk of the sporangia. Their shape is almost similar to that of Austrogramme except for the terminal cell which is glandular and not deviating from the other glandular ones. Those of T. HENNIPMAN: AUSTROGRAMME- A RE-DEFINITION 63

130 ,u

1 2 4 5 FIGURES 1-5: Paraphyses of Austrogramme. 1, A. asp len ioides (Henty & Foreman NGF 42542). 2, A. boerlageana (Buwalda 6175). 3, A. decipiens (Croo kewit s. n., L. 339327). 4, A. francii (Compton 1480, BM). 5, A. marginata (Compto n 1722, BM). vittarioides and T. f/abellivenia are essentially the same but are distinct because of the large number of glandular and non-glandular cells. In T. requiniana the paraphyses are mainly inserted on the epidermis; two different kinds were found. In BSIP 1987, 6329, 7292 (from the Solomon Islands), and LAE 53748 (from New Guinea) they are c. 375 J.lm long, straight, CIJnsist of 5- 7 cells, of which the terminal cell is curved, brownish, presumably glandular, and has rather thick cell walls, the other, non-glandular cells are thin-walled and lighter coloured. Those found in NGF 2984 1 (from New Guinea) are 500-700 pm long, curved in the upper part, consist of 5-7 cells, and are all except for the lovitermost one or two cells filled with a granular, obviously glandular substance. In T. pinnata, and T. corda ta only the terminal cell of the paraphysis seems glandular. As regards T. pinnata, over 10 brownish paraphyses consisting of 6 or 7 cells are inserted on each stalk of the sporangia in A.C. Smith 6565 (Fiji), whereas in Carr 12144 (New Guinea) few paraphyses are present on each stalk which, moreover, consist of about 11 cells. The paraphyses of T. lanceolaia (syn.: Syngramma hookeri) , and T. hosei are very different from the above as they are exclusively seated on the stalk of the sporangia, generally one to each stalk, and consist of only 2-3 elongate cells of which the terminal cell is glandular. Much variation of the paraphyses also occurs in Syngramma. Paraphyses are abse'nt in S. quinata, whi 1st those of S. a/ism ifolia, and S. luzonica appear to be mainly inserted on the stalk of the sporangium, generally one to each stalk. They consist of a somewhat swollen terminal cell which was often found surrounded by a ±spherical or pear-shaped yellowish to reddish glandular substance, and 2-:-5 hyaline cells. The paraphyses of S. borneensis are inserted in groups of 3-6 on the stalk of the sporangia, and are large, brownish, consisting of up to 10 cells of wh ich the terminal cell only seems glandular.

Spores The spores of Austrogramme, Ta enitis, and Syngramma are trilete (figs 1-3). Spores of Austrogramme have a ± circular equatorial outline with a diameter of ± 35-40 pm. In equatorial view the distal part is ± globose, the proximal part ± depressed globose. The proximal side shows a labrum and a shorter suture. The outer surface is microverrucate with superimposed more or less distinct (clusters of) short irregular excrescences. Spores of Ta enitis corda ta, T. blechniodes, T. flabellivenis, T. in terrupta, T. marginata, T. requiniana, and T. vittarioides are characterised by a conspicuous 64 FERN GAZETTE: VOLUME 11 PART 2 & 3 (19751

FIGURES 6-15: Leaf-shape and sporangia! disposition in Austrogramme. 6, 7, A. aspjenioides (Henty ·& Foreman NGF 42542). 8, 9, A. boeifageana (Kornas$i 1241). 10, 11, A. decipiens (McGillivrav 23, BM. 12, 13, A. f�ancii (Compton 1480, BM) 14, 15, A. marginata (Franc 1424). Figs 6, 8, 10, 12, 14, x Y.;figs 7, 9, 11, 13, 15: x 1%. HENNIPMAN: AUSTROGRAMME- A RE-DE FINITION 65

FIGURE 16: Spores of Austrogramme and Syngramma: a, b, A. francii (Le Rat 15, P); c, A. de cipie ns (Franc 336); d, A. boerlageana (Buwalda 6210); e, (A.· asplenioides (Henty & Foreman NGF 42542); g, h, S. quinata (Brass 3334 ); (a, c-e,g: x c. 1 000; figs b, f, h: x c. 5000). cingulum (equatorial flange) which is sometimes interrupted or reduced at the angles. a 45-70 The equ torial outline is (sub) triangular, its largest width f'm. In equatorial view the proximal part is more or less flattened or depressed globose, whereas the 66 FERN GAZE TIE : VOLUME 11 PART 2 & 3 (1975)

FIGURE 17 : Spores of Syngramma and Taenitis. a, S. luzonica (Edano PNH 40526); b, c, d, T. interruf)ta (Schiffner s.n., L. 942. 123-124), e, f, T. requiniana (Coode & Kati k NGF 29841 ), a-e: x c. 1000, f: x c. 2500). distal part is globose or only occasionally depressed. The proximal part of the spores of T. corda ta shows tori aiorig the suture, the distal part has irregularly shaped, large, raised ridges; the surface is microverrucate, bearing spaced irregularly shaped granules of varying size all over the spore. The spores of the other Ta enitis species mentioned above are all more or less similar and of the same shape as those of T. corda ta, but the proximal part without HENNIPMAN: AUSTROGRAMME- A RE-DEFINITI ON 67

F IGURE 18: Spores of Taenitis; a-d, T. cordata (Kostennans 5944 ); e, T. pinnata (Carr 12144 ); f-j, aborted spores of T. lanceolata (Brass 12908); a-c, e, f, h, i: x c. 1000; d, j : X C. 2500; g: X C. 5000. 68 FERN GAZETTE: VOLUME 11 PART 2 & 3 (1975) tori along the suture and the surface of a different shape. The upper and lower surface of the cingulum (and sma ll adjoining parts) are microverrucate, laterally densely covered with irregularly shaped baculae and granules; the outside otherwise densely packed with ± globose raised ridges or ± globose clusters of variously shapped excrescences. Spores of Ta enitis hosei, T. /anceolata, and T. pinnata lack a cingulum. Those of T. pinnata show irregular, rather labyrinth-like ridges - but are in other respects quite like the spores of T. interrupta. . Spores of T. lanceolata were found to be aborted in several specimens, th eir outside rather differing in shape. Some spores lack a suture on their proximal side, though a weak indication of ridge-like tori may be present; otherwise th e outside consists of somewhat reticulately arranged, rounded ridges. The surface of the spore is microverrucate, that of the ridges ± smooth. Other spores have a distinct suture and lack tori, their outside shows irregular reticulately arranged ridges of which the surface is densely packed with baculae and granules. Spores of Syngramma have a± circular to (sub)triangular equatorial outline with a diameter of 40-50 Jlm. Those of S. alismifolia and S. quinata are very similar in shape to the spores of Austrogramme. The proximal part bears a labrum and a shorter suture. The surface is microverrucate, and shows irregular, more or less densely placed, ridges. Spores of S. luzonica have a traingular equatorial outline, and have the distal and proximal part (strongly) depressed globose in equatorial view. The proximal part lacks a labrum, showing a long suture only. The surface is either covered with a dense felt of short excrescence throughout, or shows± smooth parts.

DISCUSS ION The combination of leaf-shape, venation pattern, and disposition of the sporangia easily distinguish the genus Austrogramme from the genera Ta enitis and Syngramma. The petiole ana�omy of these three genera differs from that found in Coniogramme (Walker 1968). · The shape of the glandular paraphyses of part of Austrogramme resembles �trongly that found in part of Ta enitis. The glandular substance was found to be inside the cell. In other species of Ta enitis different kinds of paraphyses were fou nd. The paraphyses of Syngramma, if present at all, likewise show much variation. They apparently have a single terminal glandular cell surrounded by glandular substances. Variation in paraphyses of Ta enitis and Syngramma has been noticed also by Walker (1968). The spores of Austrogramme are proportionally small. Thei r shape resembles that of some Syngramma species. The information of the spores of Austrogramme (A sp leniopsis} decipiens given by Erdtman (1957) and Erdtman & Sorsa (1971 ) is rather poor. S. a/ismifolia and S. (Craspedodictyum} quinata and S. (Craspedodic tyum} schlechteri (Erdtma n & Sorsa 1971: pi 5, fig 1) have very similar spores which su-pports the inclusion of Craspedodictyum Copeland in Syngramma as proposed by Holttum (1954). Spores of Ta enitis are very different from those of Austrogram me and Syn­ gramma, and are of two kinds either with or without a cingulum. These differences in spores are not correlated with differences in paraphyses. In Ta enitis the cingulum was first reported by Tardieu-Biot (1963) for T. ("Schizoloma"} cordata. From the publications of Erdtman (1957) and Nayar & Diwi (1966, 1967), it becomes clear that a cingulum is also found in the spores of some other genera of the Pteridaceae sensu HENNIPMAN: AUSTR O GRAMME- A RE-DEFIN ITION 69

Copeland. The occurrence of a cingulum also in Taenitis (and its American relative Syngrammatopsis - Hennipman, unpublished), seems of interest in view of the ambiguous systematic ppsition of Ta enitis. Mickel ( 1973) related Taenitis to th'e lindsaoid ferns, whereas Holttum (1960) thought Ta enitis more related to Syngramma than to Lindsaea. In both the Lindsaeoid ferns and Syngramma a cingulum is not known, however. A study of the spores seems promising when considering the relationships within Copeland's Pteridaceae. The chromosome numbe r of Austrogramme decipiens is n = c. 58 (Brownlie ex Fabbri 1963). That of Syngramma is n := 58 in S. (Craspedodic tyum) grandis (Walker 1968) and in S. alismifolia (Holttum 1960). whereas it is n = 116 in S. (Craspedodictyum) quinata (Holttum 1960). These data suggest a base number of x = 29 for both Syngramma (Walker 1968) and Austrogramme. Chromosome numbers of Ta enitis show more variation. T. blechnoides has n = 44 (Holttum 196.0) and n = 110 (Walker 1968) which point to a base number of x = 22. T. (Syngramma) lanceolata has n = 114 (Walker 1968). Differences in chromosome number in Ta enitis are associated with differences in th e spores (presence or absence of a cingulum). But apart from the cingulum, the spores of the Ta enitis species are very similar. The properties of the spores, and the chromosome number indicate a relationship between Austrogramme and Syngramma. The relationship of Austrogramme to Ta enitis seems less obvious, also_ in view of the variation of the paraphyses in the three genera studied. Further, the present study supports the emendations of Syngramma and Ta enitis as proposed by Holttum.

SYSTEMATIC TREATMENT

· Austrogramme Fournier, Ann. Se. Nat. V, 18 :278. (1873). Lllctotype (Copeland, Gen. Fil., 1947: 56): A. marginata (Mettenius) Fournier. Aspleniopsis Mettenius ex Kuhn, Ch aetopt.: 325 (1882); Diels in Engler & Pran.tl, Nat. Pf /. Fam. 1, 4: 272 (1899); Copeland, Gen. Fit.: 77 (1947 ); Brownlie, Fl. Nouv.-Ca/ed., Pterid. : 161 (1969) - Type: Aspleniopsis decipiens (Mettenius) Kuhn. Gymnogramme auct. p.p., non Desvaux. Rheop teris auct., et sensu Holttum p.p., non Alston. Rhizome short-creeping or suberect,solenostelic, clothed with inconspicuous bristles as least terminally ; roots many, firm. Leaves tufted, in several rows, simple to tripinnatifid, 9-70 cm long ; petiole dark reddish-brown, near the base or throughout (A . marginata) with inconspicuous bristles, with 2 vascular strands; lamina herbaceous, ± glabrous; veins free or variously anastomosed. Sporangia situated in (orbicular to) elongate sari, variously along the veins, with 16-20 indurated cells. Spores trilete, ± spherical, diameter of equatorial outline± 35-40 pm, whitish, thin-walled, the outer su rface± smooth (when seen with the light microscope) . Chromosome number: n =c. 58 (A . marginata). Distribution: Moluccas, New' Guinea; New I re land (?). New Hebi-ides, an_d New Caledonia. Habitat: Usually terrestrial or rheophytic, rarely (A . fra ncii) epiphytic. Altitude: 0- 1 000 m. Notes: 1. On some petioles of A. asplenioides and A. fra ncii I observed small buds situated just above the petiole base. In one leaf of A. asplenioides such a bud obviously had developed into an elongate rhizome bearing tufted roots and leaves terminally. 2. Austrogramme deplanchei (Baker) Fournier (bas.: Polypodium deplanchei Baker), the second species originally included in the genus, was transferred to ,Grammitis by Copeland (Gen. Fit. 1947 : 21 1). 3. Stomata of A_ustrogramme asplenioides and A. decipiens are polycytic. 70 FERN GAZETTE: VOLUME 11 PART 2 & 3 (1975)

Because of difference in shape of the leaves two sections are recognised :

Sect. 1. Austrogramme Fournier, Ann. Se. Nat. V, 18: 278 (1873). - Lectotype:A. margina ta Fournier. Leaves simple; veins free. Distribution: New Caledonia. Species 1-2.

Sect. 2. Aspleniopsis (Kuhn) Hennipman, stat. et sec t. nov. . Aspleniopsis �tten ius ex Kuhn, Ch aetopt.: 325 (1882). - Type Gymno­ gram me decipiens Mettenius. Leaves pinnate to tripinnatifid; veins free or anastomosing. Distribution : as the genu�. Species 3-5.

KEY TO SPECIES 1. Leavessimple. Veins free (sec. Austrogramme) . . 2 Leaves compound. Veins either (almost) free or regu larly anastomosing (sect. Asp/eniopsis} . . 3 2. Leaves 3!r-70 by 2.5-4 cm. Petiole 2-7 by 0.2

Sect. 1. Austrogramme 1. Austrogramme marginafa (Mett.) Fournier. - Fig. 14, 15, Ann. Se. Nat. V, 18: 27P. (1873). - Gymnogramme marginat:a Mett, Ann. Se. Nat. IV, 15: 59 (1861) Fourn ier, Bull. Soc. Bot. France 16: 423 (1869); Compton, J. Linn. Soc. Bot. 45: 449 (1922). - Syngramma marginata Diels in Engler & Prantl, Nat. Pfl. Fam. I, 4: 257 (1899); Brownlie, F/. Nouv.-Caled. Pterid.: 161 Vieillard 1642 (1969).- Type: (iso in BM, P). Leaves simple, 3!r-70 by 3.5-4 cm; petiole 2-7 cm long, 2-2.5 mm thick, lamina widest about the middle, long decurrent on the petiole, margin usually entire, sometimes the upper part irregularly crenate, apex acute to acuminate; veins free. Sori situated±b eyond the costa up to the margin, elongate, of varying length. Chromosomes: n =c. 58 (Brownlie ex Fabbri, 1963). Distribution: New Caledonia: Compton 1722 (BM); Franc 1424 IBM, P): McKee 6405 (K); Sarasin 64 (P); Vieil/ard 1642 (BM, P). Habitat: Primary forest. Terrestrial. 2. Austrogramme francii (Rosenst.) Hennipman, comb. nov.- Fig 12, 13. Syngramma fr ancii Rosenst., Fedde Rep. 9: 75 ( 1910);Co mpton, J. Linn. Soc. Bot. 45: 449 (1922); Brownlie, Fl. Nouv.-Caled., Pterid. : 160 (1969). - Type: Franc 1427 (P, not seen; iso in BM). Syngramma francii var. intermedia Rosenst., Fedde Rep. 9: 76 (1910). - Type: Le Rat 15 (P, 2 sh.). Leaves simple, 9-30 by 1-2.2 cm; petiole 5- 10 cm long, 1-1.5 mm thick; lamina widest about the middle, decurrent on the petiole, margin entire, apex rounded to acuminate; veins free. Sori situated either± beyond the costa up to the margin, or towards the margin only, (orbicular to) elongate, of varying length. Distribution: New Caledonia: Compton 1480 (B�; Franc 1427 (BM); Le Rat 15 (P). Habitat: Along streams; terrestrial and epiphytic. HENNIPMAN: AUSTROGRAMME- A RE-DEFINITION 71

Sect. 2. Aspleniopsis (Kuhn) Hennipman 3. Austrogramme asplenioides (Holttum) Hennipman,comb. nov. - Fig. 6,7. Rheopteris asplenioides Holttum, Blumea 11: 531 (1962).-Type: Kalkman BW 8528 (L). Leavesbipinna te to tripinnatifid, up to 25 cm long; petiole 5- 15 cm; lamina± triangular, 1.5-2 times as long as wide, 7-9 by 4-5 cm; pinnae 1G--20,altern ate, the basal ones shortly stalked, 2G--45 by 1Q--15 mm; pinnules 4-8, spaced,wedge-shaped, up to 17 mm long,distal margin ± entire or lobed to almost to costule; terminal segment of leaf and pinnae of± similar shape as the pinnules; veins free, occasionally with a d istal anastomosis. Sori elongate, of varying length. Distribution: New Guinea. West. Div. South NewGui nea,Subdiv. Muju,± 5 km NE. from junction lwur A. and East Digul: Kalkman BW 8528' (L). - East. We st. Sepik Distr., Telefomin subdistr., Prospect Creek near Frieda A.: Henty & Foreman NGF 42542 (L). Habitat: On rocks in stream-bed, inundated during high water. Note: Many inconspicuous dark brown bristles are found at the apex of the rhizome. They were not studied by Holttum. 4. Austrogramme boerlageana (v.A.v.R .) Hennipman,comb. nov. - Fig. 8,9. Syngramma boerlageana v.A.v.R., Bull. Dept. Agric. lnd. Neerl. 18: 19 (1908), pi. 3.­ Type: Boerlage s.n., Ambon (BO, not seen; phot. K). Leaves pinnate, up to 35 cm long; petiole 6-25 cm; lamina triangular, 7- 15 by 6- 10 cm; pinnae 5- 11, alternate or opposite,the lowermost ones deflexed,shortly stalked, asymmetrical, up to 60 by 30 mm, the acroscopic side of the base truncate to cordate, its basiscopic side narrowly acute, margin entire, gradually narrowing from the base towards the acute apex; terminal segment ± triangular, asymmetrical, up to 50 by 20 mm, with 1 or 2 unequally-sized basal lobes, gradually narrowing towards the acute apex. Veins forming a regu lar network of elongate,angu late areoles. Sori situated along part of the veins, elongate, often± curved, of varying size. Distribution: Moluccas. Ambon: Buwalda 6175 (L), 6210 (L) ; Boerlage s.n. (80). Brooks s.n. (BM) - Ceram: Kornassi 1241 (L). Habitat: Primary forest. Terrestrial. �- Austrogramme decipiens (Mett.) Hennipman,comb. nov. - Fig. 10, 11. Gymnogramme decipiens Mett., Ann. Se. Nat. IV, 15: 60 (186 1); Fournier,Bull. Soc. Bot. France 16: 422 (1869); Ann. Se. Nat. V, 18: 279 (1873); Compton, J. Linn. Soc. Bot. 45: 448 (1922). - Aspleniopsis decipiens Kuhn, Chaetopt. 325 (1882) ; Diets in Engler & Prantl, Nat. Pfl. Fam. I, 4: 272 (1899); Copeland, Gen. Fil. 77 (1947); Brownlie, Fl. Nouv.-Caled., Pterid. 161 (1969). - Lectotype: Vieillard 1648, New Caledonia(P, not seen). Gymnogramme decipiens var. parva Compton,J. Linn. Soc. Bot. 45: 449 (1922). - Type: Compton 1781, New Caledonia (BM). Leaves pinnate or bipinnatifid, 15-45 cm long; petiole 2-20 cm; lamina 2-6 times as long as wide, widest about or below the middle, 9-25 by 1.5-9 cm; pinnae 30,alternate or opposite,± sessile to shortly stalked, asymmetrical, the largest ones up to 45 by 1G--15 (-20) mm, the acroscopic side of the base broadly acute to subcordate, its basiscopic side narrowly acute,margin shallowlylobed to lobed to almost the costa,apex acute; terminal segment inconspicuous, its shape similar to that of the pinnae. Veins free. Sori (orbicular to) elongate,of varying length. Distribution: New Caledonia (common),New He brides,and possibly also in NewIr eland (Brownlie 1969). Habitat: In primaryforest, terrestrial and rheophytic. Note: The small rheophytic specimens were given varietal status by Compton. However, the size and shape of the plants seem largely determined by habitat conditions.

ACKNOWLEDGEMENTS I gratefully acknowledge the loan of pertinent specimens from the Curators of the herbaria at Kew, London, and Paris, and the technical assistance of Mrs J. H. Steur-van der Graaf, and Mr H. Kammeraat.

REFERENCES BROWN LIE, G. 1969. Flora de la Nouvelle-Caledonie et Depandances. Vol. 3 Preridophytes. Paris. COPE LA ND,E.B. 1947. Genera Filicum. Waltham. Mass. ERDTMAN, G. 1957. Pollen and sp ore morphology/Plant taxonomy. Gymnospermae, Pteridophyta, Bryophyta. Stockholm/New'York. 72 FERN GAZETTE: VOLUME 11 PART 2 & 3 (1975)

ERDTMAN, G. & SORSA. 1971. Pollen and spore morphology/Plant taxonomy. Pteridophyta. Stockholm. FABBRI, F. 1963. Primo supplemento alle tavole chromosomiche delle Pteridophyta di Alberta Chiarugi. Caryologia 16: 237-335. HOLTTUM, R.E. 1954. Flora of Malaya, Vol. 2. Fernsof Malaya. Singapore. HO L TTUM, R.E. 1960. Two problem species: Sch izoloma cordatum Gaud. and Syngramma pinnate J. Smith. Amer. Fern J. 50: 109-113. HOLTTUM, R.E. 1962. New species of Malaysian ferns. 81umea 11: 529-534. HOLTTUM, R.E. 1968. A re·definition of the fern-genus Taenitis Willd. 81umea 16: 86-95. MANTON, I. 1954. Cytological notes on one hundred of Malayan ferns. Appendix in: Holttum, R.E. Flora of Malaya, Vol. 2. Ferns of Malaya. Singapore. MICKEL, J.T. 1973. Position of and classification within the Dennstaedtiaceae, pp 135- 144 in: Jermy, A.C., Crabbe, J.A. & Thomas, B.A. (eds) . The phylogeny and classification of the ferns. London. NA YAR, B.K. & SANTHA DE VI. 1966. Spore morphology of the Pteridaceae. I. Pteridoid ferns. Grana Palyn. 6: 476-502. NAYAR, B.K. & SANTHA DEVI. 1967. Spore morphology of the Pteridaceae. 11. The Gymnogrammoid ferns. Grana Palyn. 7: 568-600. TA RDIEU-BLOT, L. 1963. Sur les spores de Lindsaeaceae et de Dennstaedtiaceae de Madagascar et des Mascareignes, etude de Palynologie appliquee a la systematique. Pollen et Sp ores 5: 69-86. WA LKER, T.G. 1968. The anatomy of some ferns of the Taenitis alliance. Proc. Linn. Soc. London 179: 279-286. FERN GAZ. 11(2 & 3) 1975 73

THE BIOGEOGRAPHY OF ENDEMISM IN THE CYATHEACEAE

R.M. TRYON Department of Biology and Gray Herbarium, Harvard University, Cambridge, Massachusetts, 02138, U.S.A. and G. J. GASTONY Department of Plant Sciences, Indiana University, Bloomington, ln<;liana,474 01, U.S.A.

ABSTRACT One of the principal features of the biogeography of the Cyatheaceae is the strong development of local endemism. Closely related local endemics often occur in the same region. They are most frequent where pronounced environmental mosaics have been present for considerable periods of time. Local endemics are fewer in ecologically more uniform regions, and in areas subjected to climatic ch13nge. Dataon local endemism in the Cyatheaceae support an hypothesis of local speciation based on the establishment and divergence of small peripheral populations. lt is proposed that local endemics are ecologically specialised. In spite of high dispersibility, they are isolated through their adaptation to small environmental areas not duplicated within the range of dispersal. The notable studies of Dr. R.E. Holttum on the systematics of the tree ferns (Cyatheaceae) merit recognition in this series of papers dedicated to him. His contributions to the systematics of this family include a revised generic classification and monographs of over 250 species of lndo-Asia, Malaysia, Australia and the Pacific islands.

INTRODUCTION The scaly Cyatheaceae form a large evolutionary group of about 500 species. The scales are abundant on the stem apex and usually on the leaves, especially toward the base of the petiole. Holttum (1963) included these species in the genera Cyathea and Cnemidaria, wh ile Tryon (1970a), recognising several evolutionary lines, treated them in the genera Sphaeropteris, Alsophi/a, Nephelea, Trichipteris, Cyathea and Cnemidaria. The Cyatheaceae are an excellent subject for biogeographic study because of th e large number of species, the wide geographic distribution of the family, and the recent systematic work that provides sound geographic information. The strong development of endemism, especially local endemism, has become apparent from an analysis of th e geographic ranges of the species. The nature of these endemics in the Cyatheaceae and the way they may have developed will be considered here. The data presented in th is paper have been primarily obtained from publications by Holttum (1963, 1964, 1965) and Tardieu-Biot (1951, 1953) on the paleotropical species, and from papers and manuscripts on the neotropical species by Barrington ( 1974), Con ant (1974). Gastony (1973), Riba (1969). Stolze (1974), Tryon (1971, 1972, 1974). and Windisch (1974). Some additional information has been obtained from other literature and collections in the Harvard University Herbaria.

BIOGEOGRAPHY The scaly Cyatheaceae have a pantropical distribution, with a few extra-tropical extensions to the south in South America, South Africa, Australia, New Zealand, and the Pacific, and with an extension to the north in· southeastern Asia to Japan and the Bonin Islands. The species are strongly concentrated in the major wet mountainous regions of the world (Fig. 1 ) . Data on the number of species occuring in regions .... '"

,;;;; . 1$ �7 I I . t:) 'V; I

, m ll z G) )> N m -i -i m .. I < 0 r c s: I m - - "11 i )> .I ll -i FIGURE 1: Generalized range of the scaly Cyatheaceae. The solid black areasare the major wet mountainous regions that are centres of "' species diversity. I!> w - CO .... !!! TRVON & GASTONV: ENDEMISM IN CVATHEACEAE 75

included in Table 1 clearly indicate these centres of species diversity: the Greater Antilles, southern Mexico and Central America, the Andes, Madagascar, Sumatra, Borneo, Philippine Islands, and New Guinea. The altitudinal occurrence of species is shown in Table 2. Species grow from sea-level to 4200 m, and within this wide range they occur most frequently from 1000 to 1500 m. There is a progressive decrease in numbers of species below 1000 m, and the number also diminishes at altitudes from 1500 to 2500 m. Species are much less frequent above 2500 m. Two Andean species of Cyathea grow at 4200 m, which is the highest altitude reported for tree ferns. Data on the number of endemics,. the per cent of endemism and the number of local endemics are given in Table 1. Endemics are confined to a geographic area where they may occupy a considerable range or a very small one. lt is important to distinguish between local endemics, those species with a limited distribution, and those endemics of wider distribution. The high dispersal capacity of fern spores requires a broader geographic definition of a local species than in the case of most seed plants. A range of about 300 miles (500 km) or less is considered a limited one in the ferns, and species with distributions of this size are regarded as local endeniics. The 223 local endemics (Table 1) represent 44% of the species of Cyatheaceae. The strong development of local endemism in the family seems paradoxical when the high dispersal capacity of the spores is considered. The pantropical distribution of the whole group and of wide ranging species (Fig. 2), especially those species occurring on distant islands, such as Sphaeropteris lunulata and S. medullaris, clearly demonstrate effective dispersal capacity. However, local endemics suggest that other factors such as

geographic isolation or ecological specialisation are involved in determinin· g the extent of their distribution. Geographic isolation may be the basis of local endemism in ferns if the distance is sufficient to prevent dispersal across extensive barriers to other areas of similar environment. Small and remote oceanic islands represent areas with strong geographic isolation. The distance between one island and another with a comparable environ­ ment may be so great that the chance of spores reaching it are slight. This is' especially true when there is a small source area with a correspondingly small species population. The local endemics on small remote islands of the Pacific represent examples of this type of geographic isolation. Local endemism in other regions may not be based on geographic isolation. Both local endemics and more widely distributed species occur in the wet forested mountains of the Andean area and of New Guinea. Wide ranging species are distributed along the Andes from Colombia south to Peru or Bolivia, and in New Guinea wide ranging species extend from the Milne Bay District west to .the Weyland Mountain region or the Arfak Mountains. Within these extensive areas suitable for tree ferns, the local endemics are effectively isolated. The basis for their isolation seems to be adaptation to a local environment that .is not duplicated in the region. An explanation of the manner by which ecologically specialised local endemics can evolve must consider the events leading to speciation within an ecological­ geographic framework. An ecological variant, generated from a population gene pool, that is adapted to a somewhat different environment in a peripheral area (one within easy dispersal distance) may become established there as a small population. This population will have unusual opportunities for speciation, as compared to the original one, through small population phenomena such as inbreeding and genetic drift that can rapidly alter the genetic basis for selection. In time the new population will become more closely adapted to the new local environment. A distant new species of limited range and narrow ecological adaptation may evolve when morphological 76 FERN GAZETTE: VOLUME 11 PART 2 & 3 (1975)

Numb�r of Number of Percent Number of 1 2 Region specms endemics endemism local endemics

Greater Antilles 37 28 76% 22 Southern Mexico and Central America 50 30 60% 16 Andes 3 75 57 76% 25 Brazil 24 9 37% 0 South America 127 101 79% 40

NEOTR OPICS 197 196 4 110

Africa 15 13 87% 7 Madagascar 39 35 90"/o 11 Southeast Asia (excl. Malay Peninsula) 23 17 74% 4 Sumatra 34 15 44% 9 Borneo 31 14 45% 10 Philippines 37 26 70% 13 New Guinea 79 58 73% 36 Malaysia (incl. Malay Peninsula) 186 175 94% 79 Pacific (east of Australia and New Guinea) 41 38 93% 18

PALEOTROPICS 314 313 113

TOTAL 510 223

1 Geographic subspecies and varieties are included as species. 2 With a range of about 300 miles (500 km) or less in extent. 3 State of Lara, Venezuela to Colombia, south through Bolivia to the Province of Salta, Argentina. 4 Alsophila capensis (L.f.) J. Sm. is in Africa and Brazil.

TABLE 1: Species and endem ics of Cyatheaceae in selected regions.

divergence is linked with ecological specialisation. A local endemic, derived from a more widely distributed species, may in turn give rise to another species by the same process. Such a "second generation" endemic will be even more oloec gically specialised and geographically limited as its genetic base becomes further restricted. Continuation of this mode of speciation will constantly increase the number of local endemics in a region providing that the climate remains equable. Such endemics would be highly vulnerable to extinction with a significant climatic change. The data on local endemism in the Cyatheaceae will be reviewed for its consistency with this hypothesis on the origin of local endemics. Local endemics should be most numerous in topographically diverse terrain where a pronounced environmental mosaic provides many small areas that may be occupied by peripheral populations. The data in Table 1 show the concentration of 25 local endemics in the And!!an area and 36 local endemics in -New Guinea. In contrast, much larger regions having less topographic diversity have few local endemics. For example, there are only 7 in Africa and there are non in Brazil. Related species should occur in areas within easy dispersal range of spores if they are derived by speciation of peripheral populations. All of the Greater Antilles are -i ]I -< 0 z I \. • � � r1 L� : � C) I �-i 0 z -< m z c m V � I (/J s:: z 0 -< > ··-····- ...... -i ..� .. :I: m

�m > m

I ,, \;

FIGURE 2: Widely ranging species of Cyatheaceae: 1. Alsophila manniana. (Hook.) Tryon, 2. A. dregei (Kze.l Tryon, 3. A. gigantea Hook:, 4. Sphaeropteris glauca (BI.) Tryon [Cyathea contaminans (Hook.) Copel.], 5. S. lunulata (Forst.l Tryon, 6. S. medullsris (Fom.l Bernh., 7. Trichipteris microdonta (D.esv.) Tryon, 8. Nephelea cuspidate (Kze.) ·Tryon.

...... 78 FERN GAZETTE: VOLUME 11 PART 2 & 3 (1975)

sufficiently close so that there is no ·significant barrier of distance for spore dispersal between them. The mountainous areas of adjacent islands are about 150 miles (240 km) apart. Tryon (1970b) presents evidence that distances greater than 500 miles are required for isolation of an insular fern flora. In A/sophila there is a total of 9 species in the Greater Antilles, and in Nephelea there are also 9 species; one of them with three varieties. Seven of the species of A/sophi/a and all 11 of the taxa of Nephe/ea are endemic td single islands. In each of these genera the species are most closely related to others in the Greater Antilles. Thus these local endemics must have originated ' within the topographically diverse Greater Antilles and be maintained as endemics by environmental specialisation.

Altitude Number of species Altitude Number of species

4000 m 2 2000 m 97 3500 m 8 1500 m 141 3000 m 22 1000 m 164 2500 m 52 500 m 98 100 m 51

TABLE 2: Altitudinal occurrence of 391 species ot Cyatheaceae. Species that occur only at altitudes between those listed are not included.

Altitudinal zones New Guinea Andes 1 36 local 43 other 25 local 50 other (in meters) endemics species endemics species

400 Q-4 2 00 0 0 0 2 350 Q-3900 2 6 0 4 300 Q-3400 7 10 2 9 250 Q-2900 9 14 3 18 Above 2400 17 18 4 19 200 Q-2400 12 19 8 30 150Q-1900 9 21 8 35 100Q-1400 8 21 4 39 50 Q-900 4 18 6 28 Q-1 00 2 11 5 12

As delimited in Table 1.

TABLE 3: Occurrence of local endem ics and other speciesof Cyatheaceae in altitud inal zones in New Guinea and in the Andes.

Local endemics would be expected to become extinct under environmental change. The mountainous regions of the Anclean area and of New Guinea have had contrasting climatic histories. In the Andes, from Colombia to Bolivia (ea 10°N. Lat. to 20°S. Lat.), there are extensive areas above 3500 m and many mountains rise above 5000 m. In New Guinea, extending from the equator to 10°S. Lat., the area above 3500 m is much smaller and there are only a few peaks above 5000 m. Although information on the effects of Pleistocene climates in the two regions is limited, the influence must have been more extensive and severe in the Andes than in New Guinea. In both regions local endemics occur from sea level to above 3000 m (Table 3). Seventeen local endemics occur in New Guinea above 2400 m, and there are 19 species TRYON & GASTONY: ENDEMISM IN CYATHEACEAE 79

that are more wide ranging. In the Andes there are 4 local endemics above 2400 m and 19 more widely distributed species. The number of species at the higher altitudes that

have moderate to wide ranges is similar in both regions, but the Andes· have a reduced number of local endemics in comparison to New Guinea. Local endemics can be derived by environmental restriction of a fo rmer larger range, as well as by local speciation. An interpretation of the local endemism as relict endemism, however, is not supported by the data on the species of the Andes and New Guinea. The occurrence of local endemics along with more widely distributed species throughout a wide altitudinal range (Table 3) is not a relictual pattern. In the Andes the local endemics are notably fewer at the higher altitudes where relict endemism should be pronounced.

ACKNOWLEDGErvENTS We are indebted to Dr. Alice F. Tryon for stimulating discussions of tree fern geography and critical comments on the manuscript. Data on neotropical species were obtained from studies supported by National Science Fou ndation Grant GB31170 to Rolla M. Tryon and Alice F. Tryon.

REFERENCES BARRINGTON, D.S. 1974. A revision of Trichipteris (Cyatheaceae). Ph .D Thesis. Harvard University. CON ANT, D.S. 1974. The American species of Alsophila. (MS ). GASTONY, G.J. 1973. A revision of the fern genus Nephelea. Contrib . Gray Herb. 203: 81-148. HOLTTUM, R.E. 1963. Cyatheaceae, in Flora Malesiana, ed. Van Steenis. ser. //; 1 (2) : 65-176. Groningen. HOLTTUM, R.E. 1964. The tree-ferns of the genus Cyathea in Australasia and tire Pacific. Blumea 12: 241-274. . HOLTT UM, R.E. 1965. Tree-ferns of the genus Cyathea Sm. in Asia (excluding Malaysia). Kew Bull. 19: 463-487. RIBA, R. 1969. Revision monografica del complejo Alsoph ila Swartziana Martius (Cyatheaceae). Ann. lnstit. Bioi. Univ. Nac. Autom. Mexico 38, ser. Bot. 1: 61-100. "1967". STOLZE, R.G. 1975. A taxonomic revision of the genus Cnemidaria (Cyatheaceae). Fieldiana Bot. 37: 1-98. e TARDIEU·BLOT, M. 1951. 4 Famille-Cyatheacees, in Flare de Madagascar et des Comores, ed. Humbert. Paris. TAR DIEU-BLOT, M. 1953. Les Pteridophytes de I'Afrique intertropicale Francaise. Mem. lnstit. Franc. Afrique Noir. 28 : 1-241. TRYON, R. 1970a. The classification of the Cyatheaceae. Contrib. Gray Herb. 200: 1-53. TRYON, R. 1970b. Development and evolution of fern floras or oceanic islands. Biotropica 2: 76-84. TRYON,R. 1971. The American tree ferns allied to Sp haeropteris horr/da. Rhodora 73: 1-19. TRYON, R. 1972. Taxonomic fern notes, VI. New species of American Cyatheaceae. Rhodora 74: 441-450. lRYON, R. 1974. A revision of the American genus Cyathea. (MS). WI NDISCH, P.G. 1974: The systematics of the group cifSp haeropteris hirsuta (Cyatheaceael. Mem. New York Bot. Gird. (in press ).

FERN GAZ. 11(2 & 3) 1975 81

LUNATHYRIUM IN THE AZORES

W.A. SLEDGE The University, Leeds LS2 9JT

ABSTRACT Evidence is presented that the plant known in the Azores as Diplazium allorge i, long considered to be an introduction. belongs to the Lunathyrium japonicum species-complex of eastern Asia, and is inseparable from Diplazium /asiopteris Kunze. Its great increase and spread to many islands of the Azores within a relatively short period is wholly consistent with its being an introduced species.

INTRODUCTION AND TAXONOMY The fern currently known as Diplazium a/lorgei Tard.-81. was first recorded from the Azores by Christ (1907). He correctly recognised it as a Diplazium of the D. jap onicum group and identified it, after comparison with Nilgiri examples, as D. Jasiop teris Kunze. Christ's description and comments on this fern are given in a supplementary note on introduced species in which Pityrogramma calomelanos and Adian tum hispidulum are mentioned as other introductions. That the Diplazium was indeed introduced, either intentionally or accidentally, into its original locality at Feteiros (Sao Miguel) seems clear from the quoted remarks of Dr Carreiro whose collections formed the basis of Christ's papers. The quotation refers to the locality as a planted woodland on the estate of a rich landowner who was an enthusiastic gardener and who had imported many exotic trees, shrubs and ferns for stocking his gardens. When Mme Tardieu-Biot (1938) described D. allorgei she was in 'no doubt that it was the same fern as that previously reported by Christ and since the collections of M & M me Allorge on which her description was based, like those of Dr Carreito, also came from an artificial habitat - a Cryptomeria woodland - she was equally opposed to considering it an indigenous species. As regards the distinguishing features of D. a/lorgei, its author, whilst agreeing that it was "tres voisine du D. lasiopteris", considered that it was specifically separable by "son petiole et son rachis noir, son contour plus largement deltoide, ses lobes separes par de plus larges sinus, ses sores plus longs et surtout, comme le fait remarquer Christensen (in litt.) son indusie ciliee-frangee". The author's description of D. a/lorgei reads "stipite cum rachi ochraceo vel nigro": Christ's description makes no mention of black colouration but correctly refers to the stipes as "ochraceo". His description is applicable to all Azores specimens seen by me, none of which could properly be described as having black stipes,though a variable degree of darkening in the lower part is not uncommon. This applies also to D. /a siopteris. I am unable to find any significant or constant difference between Azores specimens and Indian specimens in the supposed distinctions based on the width of the sinus between the lobes of the lamina or in the lengths of the sori, whilst a ciliate-fimbriate indusium is also typical of D. lasiop teris and indeed of the D. japonicum group as a whole. As regards the more deltoid outline of the fronds in Azores plants, a distinction to which both Christ and Tardieu-Biot refer, this is true of the original gatherings, which could well have been the product of one imported genotype, but it is certainly not true of many subsequent gatherings which show no difference from D. lasiopteris or D. japonicum in this respect. This is well shown in Wilmanns and Rasbach's illustration (1973, fig. 16) which depicts a typical plant. 82 FE RN GAZETTE: VOLUME 11 PA RT 2 & 3 (1975)

None of the characters attributed to the Azores fern therefore in my opinion justify specific recognition. The extent to which some characters in the Lunathyrium japonicum species-complex are determined genetically or by environmental conditions as unknown and the range of variation between different individuals within populafions of a single species is such that I do not believe Azores plants are distinct from all other described species. As regards the correct name to be applied' to the Azores fern the choice would seem to lie with one of the following three species.

1. Lunathyrium japonicum (Thunb.) Kurata, Journ. Geobot. 9: 99 (1961 ). Asplenium japonicum Thunb., FI.Jap. 3;34 (1784). Thunberg described his species from mountains near Nagasaki. No authentic specimens are known to exist. There are none in Thunberg's herbarium and Swartz ( 1806) indicated that no specimen had been seen by him. The current interpretation of Lunathyrium japonicum sensu stric to is thus based on a traditional concept unsupported by any authentic example. In Serizawa's (1973) recent account of the L. jap onicum complex in Japan 17 taxa are recognised - seven species, four varieties and six hybrids. Three species are cited as occurring in the Nagasaki area, the species construed as L. japonicum s.s. being separated from L. petersenii largely on the basis of its dimensions according more closely with those attributed by Thunberg to his species than with the form of L. petersenii which also occurs there. The larger L. dimorphophyl/um also grows in the Nagasaki area. The dimensions attributed by Thunberg to his species were equivalent to stipes 65 cm. long and pinnae 8 x 2.7 cm. In L. petersenii the stipes are cited as 13.5 cm long and the pinnae 5.4 x 1.35 cm. Such figures imply measurements based on individual specimens. That no distinction can be drawn between them on size alone is clear from Serizawa's own specimens (kindly sent by Prof. Reichstein). Serizawa 6246 of L. japonicum s.s. from Tokyo bears two fronds with stipes 27 cm and 35 cm and the largest pinnae are 6 x 1 cm. In Serizawa 11725 of L. petersenii from Ryukyu the stipe is 63 cm long and the largest pinnae are 12 x 2 cm. lt is evident therefore that no reliance can be placed on size.

2. Lunathyrium petersenii (Kunze) H.Ohba, Sci.Rep.Yokosuka City Mus. 11: 53 (1965). Athyrium petersenii Kunze, Anai.Pterid. 24 (1837). This was described from specimens "ex insula prope Canton". Kunze states after his description that "There appears to be none akin to A. petersenii unless A. japonicum not seen by me and not recently recognised", adding that Athyrium japonicum wou Id appear to be satisfactorily distinguished by its larger size. Kunze's herbarium was destroyed during the Second World War and I have failed to trace any authentic specimen of his species in herbaria at London, Copenhagen, Leiden, Stockholm or Berlin. There are also no photographs either of this or Lunathyrium japonicum in the extensive collections of photographs of type specimens made in herbaria throughout Europe by Alston or Morton. Since no type or authentic specimens of either L. japonicum or L. petersenii can now be traced I consider it unproved that th e plants to which these names were original ly applied belonged to distinct species. Christensen (1934) had already come to the same conclusion. The two species which Japanese botanists currently recognise under these names therefore require typification by the selection and designation of W.A. SLEDGE: LUNATHYRIUM IN THE AZORES 83 representative neotypes.

3. Diplazium lasiopteris Kunze, Linnaea 17: 568 (1843). Th is was described from specimens sent to the Berlin Botanic Garden from the Birmingham Botanic Garden in England. Later, Kunze (1851) referred specimens from the Nilgiri Hills, S. India, to D. lasiopteris. There is an authentic Kunze specimen in Herb. Brit. Mus. From specimens collected by Zollinger in Java, Kunze (1848) described another species under the name Allantodia deflexa. The type of th is is at Geneva and Morton's photograph of it (kindly sent by D.B. Lellinger) shows a small plant with creeping rhizome and fronds up to 14 cm in length, the stipes and rhachises of which are hairy throughout and beset with scales. When Christensen (1906) treated Al/an todia deflexa as synonymous with Diplazium petersenii he also included Diplazium lasiopteris as another synonym, but it seems to me that the very scaly and hairy axes of the Java plant accord better with small examples of S. Indian and Ceylon D. lasiopteris than with Japanese or Canton (or Hong Kong) plants. I have previously pointed out (Sledge 1962 ) that some plants from Java seem inseparable from D. lasiop teris. Although Christ named the Azores fern Diplazium lasiopteris he had previously (1902) followed Baker (1874) in treating D. lasiopteiis as synonymous with the earlier published Asplenium petersenii and Christensen (1906) also adopted th is opinion. Later however Christensen (1934) rejected this view, unequivocally citing Diplazium petersenii as synonymous with Thunberg's Asplenium jap onicum. The grounds for th is change of view one must suppose to have been based on good reasons though I have not traced any published statement by Christensen in justification of his altered opinion. In my account (Sledge 1962) of the athyrioid ferns of Ceylon, it was pointed out that the Ceylon Diplazium th waitesii is the same as the Nilgiri D. lasiop teris and the latte r name was used rather than the older D. petersenii partly on geqgraphical grounds and partly on account of Christensen's statement as to the latter being equivalent to D. japonicum s. s. rather than to D. lasiop teris.

CYTOLOGY The Lunathyrium japonicum species-complex includes diploid, tetraploid, pentaploid and hexaploid races based on a monoploid n = 20 as against n = 21.in Diplazium. Diploid, tetraploid and pent�ploid races have all been recorded from near Darjeeling by Bir (1961 ), who also (1965) recorded tetraploid plants from S. India. Manton (1954) records a tetraploid "D. japonicum" from Malaya and {1954) a hexaploid "D. th wa itesii" from Ceylon. Serizawa (in litt.ad Reichstein) reports Japanese counts on Luna thyrium jap onicum as hexaploid and on L. petersenii as diploid and tetraploid. The Azores fern is a tetraploid. Bir (1961) noted that in diploid plants of "jap onicum" the fronds and spores were smaller than in tetraploid "jap onicum ", but Serizawa (1973) makes no mention of differences between diploid and tetraploid L. petersenii. lt is clearly not possible at present to relate these chromosome counts to any clear-cut morphological differences betWeen the plants which yielded them. Cytotaxonomic and experimental studies are urgently needed for a proper understanding of the taxa concerned since we do not know the extent to which characters are environmentally or genetically controlled. Until such an investigation has been made, identifications should conform with currently accepted names with the recognition that their validity is provisional only and subject to confirmation or rejection following experimental studies. 84 �='ERNGA ZETTE: VOLUME 11 PART 2 & 3 (1975)

IDENTI F ICATION As regards the identity of the Azores fern, for the reasons stated above I reject the view that D. allorgei is a distinct species. Since no type specimens of L. jap onicum or L. petersenii are available for comparison we can only compare Azores specimens with gatherings from the geographical areas whence the three species concerned were originally described. Such a comparison shows that Azores specimens match Kunze's own specimen of D. lasiopteris in the Brit.Mus. and other collections from South India and Ceylon more closely than those from Japan, Canton and Hong Kong. Eastern Asiatic gatherings, wh ilst far from uniform in size or in the clothing of the stipes and rhachises are less scaly along the frond axes and much more weakly clothed with multicellular hairs. I consider therefore that Christ's identification of the Azores fern was correct · and that following the separation of Lunathyrium from Diplazium the proper name for it should be Lunathyrium lasiopteris (Kunze) comb. nov. Basionym: Dip/azium lasiopteris Kunze in Linnaea 17: 568 (1843). Syn.: Diplazium allorgei Tardieu-Biot in Notui.Syst., Paris 7: 150 (1938).

DISCUSSION Although the source of the plants first introduced into Sao Miguel is unknown, probability is strongly in favour of an Indian origin. Portugal at that time had several territorial possessions around the coasts of India and trade between the twocou ntries flourished. lt is significant also that the two other ferns mentioned by Christ as introduced species are both widespread in India. Other species introduced into Azores which are widespread in India include Cyrtomium fa/catum (L.f.) C.Presl, Diplazium esculentum (Retz.) Sw., Pteris vittata L. and P. multifida Poir., though these are too widespread in south-east Asia for any conclusion to be drawn as to the likely sources of their introduction. The number of introduced Japanese and east·Asiatic species, both flowering plants and ferns, is very few; in Dansereau's (1961) geographical analysis of Macaronesian Pteridophytes Pteris multifida is the only species cited from oriental Asia. Jermy's (1964) attribution of a possible South American origin for D. allorgei is incomprehensible, since regardless of whether Azores plants are, or are not taxonomically distinct, there can be no question as to their belonging to L. jap onicum sensu lata and this is unknown in South America. When Mme. Tardieu-Biot described D. allorgei she recognised that it was probably an introduction. Subsequent students of Azores ferns (e.g. Palhinha 1943, 1966 and Vasconcellos 1968) have seen no reason to question its adventive origin. Since this fern was first recorded on Sao Miguel it has been found in six other islands, viz. Faial, Flares, Santa Maria, Sao Jorge, Pica and Terceira (Ormonde & Paiva 1973). lt is now "very common" on Sao Miguel and on Faial (Reichstein, pers.comm.), and Ormonde has sent me details of ten different localities on Sao Miguel and Terceira where he has collected specimens. Wilmanns and Rasbach (1973) also state that it is now a common plant on Sao Miguel and they indicate 13 different localities for it on their map of th e island. This great increase and spread to other islands within a relatively short period is wholly consistent with its being an introduced species (as was recognised by Christ) and is incompatible with the endemic status which Jermy (1970) has claimed for it. Indeed any enquiry into the history of this fern in the Azores or consultation of Christ's paper would surely have precluded either of JermV:s claims being made and it is unfortunate that his statement as to endemism has now received wide circulation through quotation in the Atlas to Flora Europaea. Christ's original opinion that this fern is an introduced plant specifically indistinguishable from Nilgiri W.A. SLEDGE: LUNATHYRIUM IN THE AZORES 85

plants is in my view correct and I consider that subsequent statements relating to its identity and origin have raised a superstructure of errors on a foundation of truth.

ACKNOWLEDGEMENTS My grateful thanks are due to Jorge Paiva and J. Ormonde, lnstituto Botanico, Coimbra, who first drew my attention to this matter and supplied specimens. Although the specific status of Diplazium allorgei had been accepted by other students of Azores ferns they had already formed the opinion that it does differ significantly from Indian specimens but they asked. me to look further into the problem of its status and correct name. They are also in no doubt that this fern is an introduced and not an indigenous species. I wish to thank also Prof. T. Reichstein for the loan of Japanese specimens sent by Serizawa, Or J. D. Lovis for cytological examination of Azores plants kilidly supplied by A.C. Jermy and T. Reichstein, and J.A. Crabbe for much time and helpful assistance given.

REFER ENCES BAKER, J.G. 1874. In Hooker, W.J. and Baker, J.G., Synopsis Fi/icum ed.2, 235. London. BIR,S.S. 1961. Cytomorphological observations on Diplazium japonicum (Thbg.) Bedd. Research Bull. (N.S.) Panjab Univ. 12: 119-133. 81R, S.S. 1965. Chromosome numbers in some ferns from Kodaikanal, South India. Caryo/ogia 18: 107-115. CHR IST, H. 1902. Filices Bodi nierianae, determinees et decrites. Bull. Acad. Geogr. Bot. (M ans). . �5. CHR IST, H. 1907. Filices Azoricae leg. Dr. Bruno Carriero. Bull. Acad. Geogr. Bot. (Mans). 152- 160. CHR ISTENSEN,C. 1906. Index Filicum. Hagerup. Hafniae. CHR ISTENSEN, 1934. 1ndex Filicum: Supplementum tertium pro annis 1917- 1933. Hafniae. DANSER EAU, P. 1961. Etudes Macaronesiennes 1. Geographie des Cryptogames Vascu laires. Agron. Lusit. 23: 151-181. JER MY,A.C. 1964. Diplazium, in Tutin,T.G. et al. Flora Europaea Vol. 1. Cambridge.' JERMY, A.C. 1970. In Ward, C.M. Pteridophytes of Flares (Acores) : A surveywith bibliography. Brit. Fern Gaz. 10: 119-126. KUNZE, G. 1848. Filices Javae Zollingerianas aliasque ex herbaria Moricandiano observationes. . Bot. Zeit. 1848: 191. KUNZE, G. 1851. Filices Nilagiricae. Linnaea 24: 270. 1954. MANTON, I. In Manton, I. and Sledge, W.A. Observationson the cytology and' taxonomy of the Pteridophyte flora of Ceylon. Phi/. Trans. R. Soc. (Ser. B) 238: 127- 1 85. MANTON, I. 1954. Appendix in Holttum, R.E., Flora of Malaya Vol. 2 Ferns. Singapore. OR MONDE, J. & PA IVA, J.A.R . 1973. Additiones et Adnotationes Florae Azoricae - 1. An. Soc.

Brot. 39: 39-52. . • PA LHINHA, R.T. 1943. Pteridofitos do Arquipelago do s Acores. Bot. Soc. Brot. se'r. 2, 17: 21 5-249. PALHINHA, R.T. 1966. Catalago das Plantas Vasculares dos Acores. Texto revisto por A.R . Pinto da Silva. Lisbon. SER I ZAWA, S. 1973. Lunathyrium japonicum complex in Japan,Ry ukyu and Taiwan. Sci. Rep. Ta kao Mus. Nat. Hist. 5: 1-28. SLEDGE, W.A. 1962. The Athyrioid ferns of Ceylon. Bull. Brit. Mus. (Nat. Hist.) Bot. 2: 275-323. SWARTZ,0. 1806. Synopsis filicum. Biblio. Nov. Acad. Kiliae. TAR DIEU·B LOT,M.L . 1938. Sur un Diplazium des Acores. Notul. Syst. (Paris) 7: 149- 150. VASCONCE LLOS, J.C. 1968. Pteridofi tas de Portugal continental e llhas Adjacentes. Lisbon. WILMANNS, 0. & RASBACH,H. 1973. Observationson the pteridophytes of Sao Miguel, Acores. Brit. Fern Gaz. 10: 315-329.

FERN GAZ. 11(2 & 3) 1975 87

THE GAMETOP HYTE OF CHINGIA PSE UDOFEROX

LENETTE R. ATKINSON Amherst College, Amherst, Mass.01002, U.S.A.

. ABSTR ACT The gametophyte of Chingia pseudoferox, while conforming to the general thelypteroid gametophytic pattern, de.v iates from it by the frequent presence on margin and surface of a multicellular structure, tentatively interpreted as a hair and characterized by a reddish-brown terminal cell and simple glandular hairs borne laterally. These structures set apart the thallus of C. pseudoferox from other known thelypteroid gametophytes and support the views of those authors who propose taxonomic segregation.

INTRODUCTION The fern Chingia pseudo ferox Holttum, as the name implies, is very like the type species, C. ferox ,( Bl.) Holttum (Holttum 1971), a large fern of the mountains, usually near streams, with an erect caudex and reduced basal pinnae. The chromosome number is thought to be 36, based on a count for C. atrospinosa by T.G. Walker and another for the Phillipine fern C. urea by M.G. Price (Holttum 1974: 14). The spores of this fern were collected in 1969 by Dr R.E. Holttum in the Cameron Highlands, Malaya, at 4000-6000 feet and sent to me under the name Cyclo sorus ferox sensu Holttum (1954 : 265) with the admonition that the Malayan plants are distinct from Blume's Java species and that this fern and its allies constitute a group not yet recognized and probably needs a new name. In 1971, Holttum segregated the group from Cyc/osoru s sensu Copeland (1947) under his new genus Chingia, named in honor of the Chinese pteridologist, R.C. Ching, whose application of the principles of Christensen ( 1913, 1920) to Old World sporopt1Ytic fern characteristics (Ch ing 1940) preceded the intensive study of Thelypteridaceae which has since taken place (Ching 1963; Morton 1963; Holttum 1969, 1971; Holttum, Senl & Mitra 1970; lwatsu ki 1962- 1965; Smith 1971 , 1973, 1974 ). Relatively few gametophytes of the large assemblage known as the thelypteroid ferns have been described. Among those I know, I find a similarly recognizable basic pattern (Atkinson 1971; Atkinson & Stokey 1973) : a rather th in cordate thallus, with colorless rhizoids, which has developed from the terminal cell of the germination filament (exceptions - Coryphopteris arthrotricha Holttum ined., Th elyp teris to tta (Thunb.) Schelpe =Cyclosorus gongylodes sensu Copeland, where the germination filament ends in a hair necessitating plate initiation from subterminal cells), marginal and simple surface hairs bearing a colorless or yellow secretion, a 4-5 cell thick midrib bu lging on the ventral side, an advanced type of antheridium appearing usually on separate thalli after or before the archegonia. Superimposed on the basic pattern, and interpreted as of more recent development, are observed other gametophytic characteristics, notably additional types of hairs which have appeared among the Old World species where, judging by these differences and by the variety of sporophytic differences (Holttum 1971 ), the thelypteroid ferns are evolving rapidly. lt seems therefore profitable to continue the search in the gametophyte for evidence of relationships which may add weight to that already expressed for the sporophyte generation. 88 FERN GAZETTE: VOLUME 11 PART 2 & 3 (1975)

OBSERVAT IONS The spore of Ch ing ia pseudoferox is black and appears minutely spiny under the light microscope (fig 1). lt measures about 40.9 (39-42.9) x 27.7 (27.3-3 1.2) )Jm. Germination takes place on agar within ten days (figs 2 & 3). The germination filament is composed usually of 4-5 but may be of as many as 9 cells (figs 4, 6). On a slightly more acid medium (pieces of crock over peat) germination and growth are slower. The filament may branch and the rhizoid is often delayed (figs 5, 9). Development of a cordate thallus occu rs along the lines of the basic pattern. The plate is initiated in the terminal cell of the germination filament by oblique walls (figs 7, 8). An apical cell, cutting off segments alternately to left and right, or right to left (fig 11) appears early and gives way to an apical meristem in our one and a half months (fig 12). Simple chlorophyllous hairs (fig 13) wh ich become glandular (-figs 14, 15) appear in about a month on the margin of the thallus and in six weeks also on the su rface. Winter cultures in a north window when days are dark (December to February) were spatulate (fig 10) but cultures grown later in the year (March to August) or under artificial light were broadly cordate with a deep narrow notch or with slightly overlapping wings (figs 19, 21 ). The outside walls of the margi nal cells are undulate (figs 17, 18). At 6 months the thallus is 4-5 cells thick (fig 39). The rhizoids are colorless. In about 3 months, 3-4 celled hairs appeared, first on the anterior margins (fig 22), later on the surface near the notch of some thalli in agar culture (figs 26, 27). These hairs are not numerous in either location. They are much larger than the simple glandu lar hairs, and the first wall separating the hair from the mother cell is oblique, joining the posterior wall of the mother cell in such a way that the hair rests not only on the margin but also partly on the face of the mother cell. The basal and terminal cell of the hair are larger than the cells in between. Chloroplasts are very sm all and soon disappear (fig 23 ). As the hair increases in size the contents of the terminal cell become granular and of reddish-brown col or. The terminal cell is easily broken and then the contents drain away (fig 24). The basal cell divides, accomodating to the expansion of the adjacent cells of the thallus proper (fig 24). The intervening cells in the majority of cases are 3 in number and one, or perhaps more, may divide (fig 29, 30). These cells may also bear laterally simple hairs with a colorless glandular secretion or with a granular reddish-brown head (fig 25). Such large multicellular branched hairs may perhaps be capable of continued growth as one example was observed of a long filament of 14 cells on the surface near the notch (fig 31). The terminal cell of th is filament was granular and reddish-brown, and two short glandular hairs were borne on the 7th and 11th cells. Antheridia appear at 2-2Y:I months on the surface of small meristematic thalli (fig 16) and from 3-6 months were observed infrequently at the posterior end of archegoniate thalli or on the wings (figs 20, 21 ). The antheridium (figs 33-35) is oval to globular with a fu nnel-shaped basal cell. On old thalli the antheridium is sometimes borne on a 1-celled stalk. Dehiscence is by a pore (fig 36). Archegonia appear in about 2 months and develop in the usual way. The mature archegonium is small in my cultures. lt shows two basal cells in longitudinal section (fig 37) and a short neck projecting at right angles from the thallus or inclined slightly toward the posterior end of the thallus. The cells of the thallus surrounding the egg as well as the lower cells of the neck divide when the archegonium is mature and together with the basal cells form a jacket of small cells about the egg (fig 38). Regeneration, a common characteristic among thelypteroid gametophytes, takes L.A. ATKINSON: THE GAMETOPHYTE OF CHINGIA PSEUDOFEAOX 89

M u 2 7 bN· h 8

J/1 �- �� ,@ 14 •• .

. IS 0 13

FIGURES 1-15. Stages in germination and development of young thallus of Chingia pseudoferox. 1: spore. 2, 3: germ ination, agar, 11 days. 4: 7-celled filament. 6: young thalll.IS, crock over peat, 1 month. 6: 9-cell filament. 7, 8: plate iniation. a-a first wall, b-b second wall. 9: germination, crock over peat. 10: spatulate thallus, agar, 1Y:. months. 11: anterior portion young thallus. 12: apical meristem, detail of figure 10. 13-16: hairs, 1Y:. months. Magnifications: 1, 15 x 500; 2- 5, 7-9, 11, 13, 14 X 320; 6, 12 X 180; 10 X 42. place under adverse growing conditions. A thallus bearing simple glandular hairs and two multicellular branched hairs on the anterior margins was torn in mo1,.1nting. On examination before discarding it sometime afterward, regenerative outgrowths were found on the surface of the thallus near the injury (fig 28). Rhizoids had not been developed on these outgrowths but two of them bore antheridia (fig 32).

DISCUSS ION AND CONCLUSIONS The gametophyte of Chingia pseudoferox differs from that of other described thelypteroid ferns in its types of hair. The stout, red-tipped, multicellular structures here called hairs, set apart the gametophyte of C. pseudoferox from other gametophytes which show the basic pattern for thelypteroid ferns. Whether this is a variation characterizing the genus cannot be said since other gametophytes in the same genus are unknown, including that of C. ferox, the type species, which C. pseudoferox resembles so closely in the sporophyte. 90 FERN GAZETTE: VOLUME 11 PART 2 & 3 (1975)

FIGURES 16-27. Mature thallus of Chingia pseudoferox. 16: antheridial thallus, agar, 2 months. 17, 18: undulating marginal cells. details of figures 16 & 21 respectively. 19: archegonial thallus, agar, 2Y. months. 20, 21: hermaphroditic thalli, 3 & 6 months respectively. 22: stages in development of marginal multicellular hairs at notch, agar 3Y. months. 23 : portion of anterior margin of thallus with simple and multicellular hairs, agar3 months. 24, 25: marginal multicelluhr marginal multicellular hairs. 26, 27: stages in development of surface multicellular hairs, agar3Y. months. Magnifications: 16 x 42; 17, 18 x 180; 19 x 24; 20, 21 x 12; 22 - 27 x 320.

The evidence for calling the red-tipped gametophytic structure in C. pseudoferox a hair and not a regenerative outgrowth, rests on their differences when they are present on the same thallus. The cells of the red-tipped structure lose their chloroplasts . or retain only very small ones, so that the subterminal cells appear colorless. Th is is a characteristic of many hairs. The red-tipped structure is borne on the anterior margin of the th allus or cushion near the notch on healthy-looking tissue and may bear other L.R. ATKINSON: THE GAMETOPHVTE OF CHINGIA PSEUDOFEROX 91

3 ' -� �

FIGURES 28-39. Regeneration and sex organs in Chingia pseudoferox. 28: injured thallus, arrows indicate regenerated areas. 29, 30: multicellular hairs, anterior margins, detail of figure 28. 31: single specimen of multicellular structure, 3 month old thallus. 32: regenerative growth, detail of figure 28. 33, 34: development of antheridium. 35: mature globular antheridium. 36: dehiscence of antheridium. 37, 38: archegonium. 39: portion of thallus, l.s., 6 rronths. Magnifications: 28 x 12; 29-3 1, 33-34, 36 x 320; 32 x 180; 35, 37, 38 x 500; 39 x 115.

hairs. The regenerative outgrowths, on the other hand, are green, look like small thalli, appear near the injured tissue and bear not only hairs but also antheridia, a characteristic of thalli. Among the thelypteroid gametophytes which I have seen, red-tipped gametophytic hairs are known in Mesophlebion crassifolium (81.) Holttum (Atkinson, in press) with a chromosome number also based on 36, but the hairs in M. crassifolium are quite different in form and abundance: they are simple or branched on a 2-3 celled stalk. They are numerous on margins and surfaces and are associated with 92 FERN GAZETTE: VOLUME 11 PART 2 & 3 (1975)

acicu lar hairs which have not been observed in C. pseudoferox. Holttum (1971: 30, 1974) finds sporophytic similarities between P/esioneuron and Chingia sufficient to suggest a relationship but the gametophytes do not wholly bear this out. Although the chromosome number may be 36, the spores spiny (light microscope). germination possible on both alkaline and acid media, marginal cells with undulating walls and an antheridium of similar form with a fu nnel-shaped basal cell, the gametophytes differ markedly in the hairs they bear. Hairs on margins and surfaces in Plesioneuron marattioides (Aiston) Holttum ar� of one type: long slender, simple with an abundant yellow secretion (Holttum; 1975). The hairs in Chingia pseudoferox , are of two types: either simple, short and stout with a colorless secretion, or, multicellular structures with a reddish-brown terminal cell and simple glandular hairs sometimes borne laterally. Holttum ( 1971 : 31, 1974) also finds sporophytic similarities between Chingia and G/aphyropteridopsis but the mature gametophytes of my cultures of C. pseudoferox and of G. erubescens (Hook.) Ching the type species of the latter (Atkinson, 1971) are not similar, although the chromosome number is thought to be based on 36 and the spores are somewhat alike. Unlike C. pseudofe rox, the margins of the mature thallus (3Y. months) of G. erubescens were irregular (although the walls of individual marginal cells were not undulate) and the thallus bore only simple glandular hairs with a yellow secretion. The large spore of C. pseudoferox suggests a polyploid condition in th is fern. Thelypteroid ferns whose spores (according to my approximate measurements) are over 40 pm in length have been tetraploid (Atkinson & Stokey 1973, Atkinson, in press): these include Jamaican plants of Th e/ypteris torresiana (Gaud.) Ching, T. to tta (Thunb.) Schelpe, T. norma/is (C. Chr.) Small, T. reticulata (L.) Proctor and the Malayan Metathe/ypteris singa/anensis (Bak.) Ching and Mesophlebion crassifolium. The spores of C. atrospinosa (C. Chr.) Holttum measure about 30.8 x 19.5prn. This suggests a diploid plant, judging by my measurements of other thelypteroid fern spores from plants known to be diploid. The gametophyte of C. pseudoferox, while conforming to the general thelypteroid gametophytic pattern deviates from it by the presence of a second type of hair not hitherto observed: a robust multicellular structure with a reddish-brown terminal cell and simple glandular hairs borne laterally. This characteristic must be looked for among other gametophytes of the Ch ingia group before its value can be assessed. The multicellular gametophytic hair borne by the C. pseudoferox thallus separates it from other known thelypteroid gametophytes but (beyond a similarity in the terminal cells of otherwise dissimilar gametophytic hairs borne by Mesophlebion crassifo/ium and by C. pseudoferox) does not suggest relationships. The gametophytic characteristics of C. pseudoferox, as interpreted here, do, however, support taxonomic segregation such as that proposed by Holttum on the basis of sporophytic characteristics and would seem to indicate a higher level of evolution among the thelypteroid species than those bearing only one type of hair.

REFERENCES Jl,TKINSON, L.R. 1971. The gametophyte of Th elypteris erubescens. Am. Fern J. 61: 183- 186. ATKINSON, L.R. 1975. The gametophyte of five Old World thelyptercrid ferns. Phytamorphology. In press. ATKINSON, L.R. & STOKEY; A.G. 1973. The gametophyte of some Jamaican thelypteroid ferns. Bot. J. Linn. Sac. 66: 23-36. CHING, R.C. 1940. The natural classification of the family "Polypodieceae". Sunyatsenia 5: 21o-268. L.R. ATKINSON: THE GAMETOPHYTE OF CHINGIA PSEUDOFEROX 93

CH ING, R.C. 1963. A re-classification of the family Thelypteridaceae from the mainland of Asia. Acta phytotax. sin. 8: 289-335. COPE LAND, E.B. 1947. Genera Filicum. Waltham, Mass. CHRISTENSEN, C. 1913. A monograph of the genus Dryopteris. Part 1. K. da nske Vidensk. Selsk. Skr. (Ser. 7) 10: 56-282. CHR ISTENSEN, C. 1920. A monograph of the genus Dryopteris. Part 2. K. da nske Vidensk. Selsk. Skr. (Ser. 8) 6: 1- 132. HO LTTUM, R.E. 1954. A revised Flora of Malaya, 11, Ferns of Malaya. Singapore. HOLTTUM, R.E. 1969. Studies in the family Thelypteridaceae. The genera Phegopteris, Pseudophegopteris, and Macrothelypteris. Blumea 17: 5-32. HOLTTUM, R.E. 1971. Studies in the fam ily Thelypteridaceae Ill. A new system of genera in the Old World. Blumea 19: 17-52. HO LTTUM, R.E. 1974. Studies in the family Thelypteridaceae. VII. The genus Ch ingia. Kalikasan, Philipp. J. Bioi. 3: 13-28. HOLTTUM, R.E. 1975. Studies in the fam ily Thelypteridaceae. VIII. The genera Mesophlebion and Plesioneuron. Blumea 22: 223-250. HOLTTUM, R.E., Sen, U, & MITRA, D. 1970. Studies in the fam ily Thelypteridaceae. 11. A comparative study in the type-species of Th elypteris Schmidel, Cyclosorus Link, and Ampelopteris Kunze. Blumea 18: 195-215. IWATSUKI, K. 1962-1965. Taxonomy of the thelypteroid ferns, with spec ial reference to the species of Japan and adjacent regions. Parts 1-4. Mem. Coli. Sci. Kyoto Univ. (B), 30: 21-51;37: 1-40; 125- 197. MORTON, C. V. 1963. The classification of Th elypteris. Am. Fern J. 53: 149-154. SMITH, A.R. 1971. Systematics of the nee-tropical species of Th elypteris Section Cyclosorus. Univ. Calif. Pubis. Bot. 59: 1-136. SM ITH, A.R. 1973. The Mexican species of Thelypteris subgenera Amauropelta and Goniop teris. Am. Fern J. 63: 116- 127. SMITH, A.R. 1974. A revised •classification of Th elypteris subgenusAmauropelta. Am. Fern J. 64: 83-95. ,I FERN GAZ. 11(2 & 3) 1975 95

TAXONOMIC NOTES ON SOME AFRICAN SP ECIES OF ELAPHOGLOSSUM

R.E.G. PICHI SERMOLLI lnstituto Botanico deii'Universit'a� 74 Borgo XX Giugno, 06 100 Perugia, Italy.

ABSTRACT The species of the group of Elap hogllJssum deckenii s.l. in continental Africa are considered. Full descriptions are giveri for E. deckenii (Kuhn I C. Olr., E. kuhnii Hieron., E. ruwenzorii Pirrotta, and E. tangsnjicense Krajina ex Pic.Ser., and a key to their identification provided.

INTRODUCTION The treatment of the African species of Elaphoglossum in. Schelpe's Reviews of tropical Afric an Pteridophyta, 1 (Schelpe 1969) has undoubtedly greatly contributed to our knowledge of the taxonomy of the African species of the genus, but some species need further consideration. This paper deals with the group of Elaphoglossum deckenii in continental Africa, and is intended as an addition to the treatment of.that

group in my account on the Elaphoglossaceae in the Adumbra· tio Florae Aethio· picae (Pichi Sermolli 1968).

GENERAL DESCR IPTION AND DISTRIBUTION OF GROUP Elaphoglossum deckenii and related species belong to the group of species with flat, strongly ciliate and variously coloured scales and with sterile fronds den5ely paleaceous on the undersurface and at the edges. The group was named'Chromatolepideae by Fee (1852 ). The circumscription of the species of this group is rather confused;un �oubtedly the species are not very sharply distinct, but the disagreement am,ong authors about the taxonomical value of the various entities is mainly due to the fact that, in distinguishing them, some pteridologists dislike taking into consideration the structure and colour of the scales, wh ich on the contrary offer us the best characteristics to distinguish one species from �nother. The study of many specimens belonging to both gerontogaean and neogaean species of the group has shown me that, besides the structure and colour of the scales of the rhizome, stripe and lamina, the best characteristics to distinguish the different species are the features of the rhizome, the size of the plant, the length of the sterile frond in comparison.with the fertile one, the length of the stipe in comparison with the blade, the degree of toughness of the stipe, the shape of the· blade, and the colour and the degree of covering of the lower and upper surfaces of the sterile blade. The group is distributed throughout the tropics, particularly in America where . � we can enumerate many spec1es, e.g. E. muscosum (Swartz) Moore, E. perelegans (Fee) Moore, E. plumosum (Fe'e) Moore, E. elegans Hieron., E. meridense (Kiotz.) Moore, E. orbignyanum (Fe'e) Moore, E. langsdorffii (Hook. & Grev.) Moore, E. laminarioides (Bory) Moore, E. chrysolepis (Fee) Alston, E. auricomum (Kunze) Moore, and E. vestitum (Schlecht. & Cham.) Moore. lt is only scarcely represented in the Old World and in the Pacific Islands. We may mention E. pa.feaceum (Hook. & Grev .) Sledge from Macaronesia, E. nilgiricum Krajina ex.Siedge and E. ceylanicum Krajina ex Sledge from India and Ceylon, and E. micans (Mett. ex Kuhn) Pic.Ser. from Hawaii Islands. 96 FERN GAZETTE: VOLUME 11 PART 2 & 3 (19757 ' As far as tropical Africa is concerned, the study of the types of all the described species from this continent and adjacent islands, and the revision of many specimens kept in the most important European herbaria have convinced me that the group is represented there by E. deckenii (Kuhn) C.Chr., E. kuhnii Hieron., E. ruwenzorii Pirotta, E. tanganjicense Krajina ex Pic.Ser., E. sp lendens (Bory ex Willd.) Brack,, E. rufidulum (Willd. ex Kuhn) C.Chr., E. leucolepis (Bak.) Krajina ex Tardieu , E. poolii (Bak.) Christ and E. multisquamos_um Bonap., the last five species being confined to Malagassia. An.other undescribed species occurs in south tropical Africa. lt has been often confused with E. kuhnii. Unfortunately I Have examined few and incomplete specimens of it and I am not in the position to describe it here. Perhaps another species is present in west tropical Africa, but I have seen only two sterile and incomplete specimens·. These latter two species are thus not included in the key. Some of the Malllgassian species of this group, for instance E. splendens and E. rufidulum, are closely related to the above-mentioned species from continental Africa. The former is endemic in Reunion, wh ile the latter occurs in Madagascar, Comoro Is. and Re'union. Both are provided with paleae of the rhizome and phyllopodium with edges and cilia concolorous with the central portion of the surface; thus they appear nearer to E. ruwenzorii and E. ta nganjicense than to the remaining two species. However they differ from them at first sight in the fertile fronds slightly longer than the steriJe ones oras long as them, in the structure of the rhizome and in the size and shapeof the fronds. Particularly good differences are found in the structure and size of the scales of the various parts of the plant.

KEY TO THE SPECIES I give below a key and comparative description of each of the known species from continental Africa. 1. Paleae of rhizome and phyllopodium with cilia and often also edgesmanif estly darker than the central portion of their surface 2 Paleae of ltlizol;ne and ph'/llopodium with edges and cilia con-colorous with the central portion of their surface . 3 2. Paleae of rh izome and phyllopodium similar, subulate to narrowly triangular, 6-9.5 mm with a light castaneous central stripe and blackish-castaneous marginal band and cilia. Fertile fronds manifestly shorter than the sterile ones, the latter usually 7Q-90 cm with blade long linear. Rhizome long-creeping, stout . . E. deckenii P!lleae of rhizome and phyllopodium different, those of rhizome very narrowly triangular, 3-4 mm, fulvous· with castaneous ciHa and sometimes margins; those of phyllopodium linear, 5-7 mm melleous to fulvous with dark castaneous broad-based spinules. Fertile fronds only slightly shorter to slightly longer than the sterile ones, the latter usually 15-22 cm with blade oblong-oblanceolate to linear-oblanceolate. Rhizome short-creeping, slender E. kuhnii 3. Paleae of stipe and midrib �lik�· in fertile and sterile fronds; paleae of phyllopodium 6.5-9 mm, thin, light castaneous throughout, up to 6 mm irregularly spread. Sterile blade very narrowly oblanceolate, long tapering downwards, slightly paleaceous, rusty-green beneath. Fertile frond with lamina manifestly rounded at the base. Rhizome short-creeping . E. ruwenzorii Paleae of stipe and. midrib· different in fertile and sterile fronds; paleae of phyllopodium 3.5-5.5 mm reddish-castaneous, thick, with shortstraight cilia; paleae of thest erile stipe up to 11 mm melleous, squarrose, with very narrow castaneous marginal strip, apical part and cilia. Sterile blade narrowly linear to narrowly linear-el liptic or linear-lanceolate, cuneate at the base, densely melleous-paleaceous beneath. Fertile frond- with lamina attenuate to cuneate, shortly decurrent. Rhizome erect . E. te nganjicense PICHI SERMOLLI: SOM!' AFRICAN SPECIJiS OF ELAPHOG LOSSUM 97

ELAPHOG LOSSUM DECKENII E; deckenii (Kuhn) C.Chr. lnd. Fil. 305 (1905). Basionyril.:Ac rostichum deckenii Kuhn, Fil. Deck. 16 (1867) .. ). 3-4000 '. Holotypus: "Kilema ad radices-- mantis Kilimandjaro (reg. Dschagga anno 1 864'� Kersten 5(B!). -- _Rhizpma long-qreeping, stout Fronds borne in two rows on the upper ·portion. of the · · rhizome. Sterile fronds (461 70-90 ( 1081 ·cm long; stipe about 2/3to 3/5 as long as the lamina, slender; lamina chartaceous, long linear, usually long attenuate ;1t thebase and abruptly acuminate to shortly cuspidate at the apex� sparsely paleaceous and rusty-green above, densely ferrugineous-paleaceous beneath. Fertile froni:ls shorter than the sterile ones (4/7 to 3/51, [with stipe shorter than that of the sterile frond and] with lamina usually cuneate to rounded at the base. Paleae of the rhizome and phyllopodium alike, 51,1bulate to narrowly triangular, 6-9.5 mm long, with a light castaneous central stripe and with -shiny, blackish-castaneous fairly broad marginal bands and cilia, the latter fairly long, more or less curved, irregular in length, direction and distance, arranged singly to 2-4 together. Paleae of the stipe of the sterile frond lanceolate, 6.5-8 mm long, fulvous, with long reddish-castaneous cilia and sometimes also with the apex and the edges irregu larly and moderately redd ish-castaneous. Paleae of · the midrib fulvous with concolorous cilia. Paleae of the surface fulvous with long concolorous cilia and with a reflexed canaliculate, hemi-infundibuliform long base; some of them large, ovate-lanc.eolate, others small irregularly stellate, the two kinds nearly in the same proportion on the lower 9Jrface, the ste llate ones predominant on the upper surface. Paleae of the stipe and midrib alike in the fertile and sterile fronds. The species occurs in the mountains (2800-3500 m) of S. Ethiopia, Kenya, Uganda, NE Congo and Tanzania. Schelpe (1969 ) records E. deckenii also in Malawi, Rhodesia and Comoro Is., but I have seen no specimen from these countries. I think that these records refer to E. tanganjicense whith certainly occur in Malawi and Comoro Is. (cf. Pichi Sermolli 1968: 238, 241 ). Schelpe's remt nks in Fl. Zambesiaca (1970: 215) that the specimens from Rhodesia and Malawi "have none or very few of the stipe scales with the black cilia usually seen in the E. African specimens".strongly support my supposition. I cannot oiscuss here E. hirtum var.giganteum Bonap. (Not. Pterid. . 14: 337; 1923) since I have not seen the type of this variety, Bequaert385 1 from Ruwenzori; it may be related.

ELAPHOGLOSSUM KUHNII E. kuhnii Hieron. Bot. Jahrb. Syst. 46: 399 (1911 ). Syntypi: "Trees. Mountains. Sierra Leone. N.E. 5.57" Barter s. n. Dupl. ex Herb.

Kew 1133 (B! Type- coli, in K! BM! ) * - Kamerun: "Bipindi. Urwald bei Station " lolodorf. 600 m. >¥ Farn. An hohen 8aumen aut den grossen Aesten. Dec. 1901 " Ze nker 2419 (81) - "Kamerun. Aedf Bacdemm bei Moliwe. VIII. 1905". Sch/echter 15790 (8!). Lectotypus: Zenker 2479 �B!)-Schelpe (1969: 35 ) designates Barter s.n. as the lectotype of the species, but his selection cannot stand since the syntype in Berlin Herbarium studied by Hieronymus is sterile, while the latter describes also the fertile fronds in the prcJtologue. Consequently one of the remaining syntypes, Zenker 2479 and·Schlechf:e; 15790, both fertile, is bet!_e_r qualified as the lectotype of the species. I select as such the former, being more complete and better preserved than the latter. Taxon. Syn.: E. camerouniense Ta rdieu in Tardieu, Nickles & Jacq. Fel., Etud.

*Hieronymus in the protologue gives "Barter n.6" as collection number and 1851 as the year of collection, but this is wrong. Probably he interpreted "N.E. 5. 57" (N iger Expedition May 18571 as "N. 6 May 1851 ". 98 FERN GAZE 1TE : VOLUME 11 PART 2 & 3 (1975)

Cameroun. 2. 92. t.3 (3-6) (1950). Holotypus: Cameroun: Makak, sur branches elevees. O:t. 1938. Jacques-Felix 233 1 (P!). Rhizome short-creeping, slender. Fronds fasdcled at the top of the rhizome. Sterile fronds ( 1 0) 15-22 (34) cm long; stipe 1/2 to 1/3 as long as the lamina, slender; lamina chartaceous, oblong-oblanceolate to linear·oblanceolate, attenuate at the base and subacute to sub-obtuse at the apex, sparsely paleaceous to nearly denudate and dark green above, densely fulvou&-melleous paleaceous beneath. Fertile fronds shorter to slightly longer than the sterile ones, with stipe usually longer than that of the sterile frond, and with lamina rounded at the apex and long decurrent at the base. Paleae of the rhizome very narrowly triangular, 3-4 mm long, fulvous with castaneous ' cilia and sometimes also with casta neous narro� marginal borders cilia fairly short, rigid, irregu lar in arrangement but mainly ascending, arranged singly or sometimes 2-3 together. Paleae of the ' phyllopodium linear, 5-7 mm long, melleous to fulvous, sometimes with a very narrow castaneous edge, spinulose-ciliate with spinules dark castaneous, short, straight to falcate, rigid, thickened at the base, irregular in length. Paleae of the stipe of the sterile frond narrowly Janceolate, 3-5 mm long, melleous with long, thin, flexuose concolorous cilia. Paleae of the lamina melleous, opaque with long concolorous cil ia and with a reflexed canaliculate hemi-infundibuliform long base; some of them large, lanceolate, others small, irregular stel late, the latter being more frequent. Paleae of the stipe and midrib alike in the fertile and sterile fronds. The species is widely distributed in the mou ntains of west tropical Africa from Sierra Leone to Cameroon at an elevation of 600-2 100 m. E. kuhnii has been recorded in Rhodesia, Malawi and Madagascar (cf. Schelpe 1969: 35"1, but as far as I know it is not present there. Neither Christensen (1932), nor Ta rdieu-Biot (1960) record this species in Madagascar. The specimens from Rhodesia mentioned by Schelpe (1970) as E. kuhnii belong to 'the undescribed species from south tropical Africa me ntioned above. Perhaps the latter occurs in Madagascar, too, but I have seen no specimen from there. As me ntioned above, two specimens from Cameroon and Fernando Po identified as E. kuhnii but with fronds of much larger size may be proved to be a distinct species; however, I cannot express any precise opinion on them since these specimens are poor and sterile.

ELAPHOG LOSSUM RUWENZOR II E. ruwenzorii Pirotta, Ann. di Bot. (Roma ) 7: 174 (1908) * Syntypi: In the original paper devoted to the descriptions of the new species collected on the western side of Ruwenzori in 1906 by A. Roccati and A. Cavalli Molinelli,_ members of the expedition of the Duca degli Abruzzi, Luigi Amedeo di Savoia, Pirotta gives only the fol lowing information on the type : "Hab. Valle Mobuku". In a later publication .on the results of the expedition (Savoia LA., 11 Ruwe nzori. Relaz. Se. 1 (Zool. Bot.). 1909 ), Pirotta, author of the account on the Pteridophytes, mentions (p. 483 ) that E. ruwenzorii is based on three specimens: "Valle Mobuku, epifita sulle Laurinee (?); Kiciuciu, scendendo da Kiciuciu a Nakitava, nella foresta di Nakitava; nella foresta di Nakitava a Kiciuciu, 8 Luglio 1906". Actually in the Tu rin herbarium among the collections of the Duca degli Abruzzi, we find three specimens identified by Pirotta as "Eiaphoglossum Ruwenzorii". They bear no collection number or collector name; they are labelled as follows : "(1) Valle Mobuku. ' Epifita sulle laurinee. Kichuchu. 8 Luglio 1906" with the field label "Kiciuciu. 8 Luglio 1906". lt consists of a sterile plant with rhizome (although incomplete). (2) , ;Valle Mobuku. Scendendo da Kichuchu a Nabitava. 9 Lugljo 1906". lt consists of seven sterile fronds·without rhizome . (3 ) "Valle Mobuku. Nella foresta di Nabitava e Kichuchu. 8 Luglio 1906". lt consists of a single sterile frond of E. ruwenzorii and five

*Volume 7 of the Annali di Botanica bears 1909 as the date of publication, but pages 1-196 of it were published on 31 August 1908. PICHI SERMOLLI: SO'v1EAFRIC AN SPECIES OF ELAPHOGLOSS UM 99 sterile fronds of E. deckenii, all without rhizome. Pirotta describes the fertile fronds in the protologue; but all the above-mentioned specimens are sterile and I have searched · in vain fo.r a fertile specimen in the Turin herbarium among the collec�ions of the Duca degli Abruzzi. On the other hand also the dupl icate in the Paris Herbarium is sterile. Probaply the above-mentioned specimens bore only one fertile frond and it has been lost. In this situation I designate as the lectotype the first of the above-mentioned . specimens being the only one provided with rhizome. Lectotypus: Ruwe nzori: "Valle Mobuku. Epifita sulle laurinee. Kichuchu. 8 Luglio 1906". Roccati & Cavalli Moline/li (Sped. Duca degli Abruzzi ) s. n. (TO!). Rhizome short-creeping, slender. Fronds fascicled on the distal part of the rhizome. Sterile fronds (20) 28-43 (51) cm long; stipe about 1/2 to 5/8 as long as the lamina, slender; lamina · chartaceous, very narrowly oblanceolate, long tapering downwards into a long attenuate base, acuminate to long cuspidate at the apex, sparsely paleaceous becoming denudate and dark green above, paleaceous and rusty-green beneath. Fertile fronds usually shorter than the sterile ones, with stipe longer than that of the sterile frond, with lamina manifestly rounded at the base. Paleae of the rhizome triangular, 3-4 mm long, blackish-castaneous, becoming gradually paler towards the base, slightly folded in their distal part, with concolorous edges and cilia, th e latter short, more or less falcate, irregular in length, direction and distance, mainly arranged sil]gly. Paleae of the phyllopodium narrowly lanceolate-triangular, 6.5-9 mm long; thin, light castaneous in the lower half, becoming gradually darker and thicker to dark castaneous in the upper part, which is a little rolled, with concolorous, thin,- flexuose, long cilia. Paleae of the stipe of the sterile fronds narrowly triangular,.3-6 mm long, spreading, light castaneous, with concolorous surface and cilia, the latter fairly long, thin, approx imate and irregular. Paleae of the lamina of two kinds: some of them narrowly lanceolate to very narrowly triangular, with long concolorous cilia and with reflexed hemi-infundibul iform short base, others sr:nall, stellate, but very scarce. Paleae of the stipe and midrib alike in the fertile and sterile fronds. E. ruwenzorii varies in the size of the fertile fronds in comparison. with the sterile ones, and in the degree of the paleaceous covering of the stipe and blade, which, however, greatly depends on the age of the frond, but the structure of the paleae and the shape of the I ami na are very characteristic and sharply distinguish this species from the others of the group. E. ruwenzorii has been collected in the mountains (2100-3000 m) of Kenya (Meru, Fort Hall, Naivashs and Nakuru districts), Uganda (Ruwenzori) and NE Congo (Ruwenzori, Mt Kahuzi, Volcan Mikeno).

ELAPHOG LOSSUM TANGANJICENSE E. tanganjicense Krajina ex'Pic. Ser. Webbia 23: 239. f.5, 6. 1968. Holotypus: Deutsch Ost Afrika (Tanganjika Terr.). Bezirk Morogoro: Ulugurugebirge. NW ea. 1480 m Nebelwald, Farn vereinzelt am B�umen u. Fclsen Kiluguru: Lussangasanga. 10. 12. 1932. Sch/ieben 307 1 (B!). 'lsotype in BM!, G!, P!, S!. . Paratypi: Pichi Sermo/li 71 18 (PIC.SER.!) from Ethiopia, Stolz 855 (B!) from Malawi, Louis 502 1 (EA!, PlC. SER.!) from Congo. Rhizome erect, short, slender. Fronds fascicled at the top of the rhizome. Sterile fronds (20) 34-50 (73) cm long; stipe about 1/� to 1/4 as long as the lamina, slender; lamina usually herbaceous to flaccid, narrowly linear to narrowly linear-elliptic or linear-lanceolate, usually cuneate at the base and acute to attenuate at the apex, sparsely paleaceous to nearly denudate and light green with midrib densely paleaceous above, densely melleou&-paleaceous beneath, margins with dense, long and squarrose paleae. Fertile fronds shorter ( 1/2 to 3/4) than tile sterile ones, with stipe longer and lamina much narrower (1/2 to 1/3) than those of the sterile frond, with lamina attenuate to cuneate, and gradually but shortly decurrent at the base. Paleae of the rhizome ovate-lanceolate, 3.2-4.3 mm long, redd ish-castaneous, shining, flat on their distal part, with cilia, the latter short, straight, irregular in length and distance. Paleae of the phyllopodium narrowly lanqeolat&, 3.&-5.5 mm long� shining, reddish-castaneous but becoming fulvous towards th�;� base, 100 FERN GAZETTE: VOLUME 11 PART 2 & 3 (1975}

111d dark castaneous towards the apex; cilia concolorous, short, slender, stiff, irregular,fairly close together. Paleae of the stipe of the sterile frond narrowly lanceOlate-triangular, 6-11 mm long, squarrose, melleous but with castaneous edges, apicpl part and cilia, the latter relatively short, rigid, mainly straight Paleae of the midrib melleous'fulvous with light castaneous apex and cilia Paleae of the lamina melleous, shining with long concolorous cilia and with reflexed, canaliculate, hemi-infundibuliform base; some of them large, lanceolate, others small, .irregularly stallate, the former beingmore frequent. Paleae of the stipe and midrib of fertile fronds different from those of the sterile fronds and darker in colour. E. tanganjicense occurs in Ethiopia, Kenya, Tanzania, Ruanda, Congo, Malawi and Comoro Is. Probably it is present also in 'Rhodesia; in fact I think that the specimens identified by Schelpe (1970: 21.5) as E.. deckenii belong here. In continental Africa it grows at an elevation of 1200-2300 m . . E. tanganjicense, E. ruwenzorii and E. deckenii occur in three different altitudinal belts, following each other, the first species growing at the lowest level, the third at the highest one. I think that E. ru wenzorii and E. deckenii may be interpreted as two altitudinal vicariants of E. tanganjicense, which is more widely distributed than theother two.

REFERENCES

Cf:I RISlENSEN,C. 1932. Pteridophytes of Madagascar Dansk Bot. Ark. 7: 162-171. F�E. A.L. 1852. Genef1! Filicum (Mem. 5): 41-43. PICHI SERMOLLI, R.E.G. 1968. Adumbratio Florae Aethiopicae: Elaphoglossaceae Webbia 23: 209-246. SCHE LPE, E.A.C.L.E., 1969. Reviews of tropical Africa Ptaridophyta 1. Contrib. Bolus. Herb. 1: 1-132. TA RDIEU-BLOT, M.L. 1960. Fl. Madag. Fam. 5. 2: 21-57. FERN G 'i>..Z. 1 1!2 & 3) 1975 · 101

OBSERVATIONS ON THE SPREAD OF THE AMERICAN FERN PI TYROGRAMMA CALOME LANOS

E.A.C.L.E. SCHELPE Bolus Herbarium; University of Cape Town

ABSTRACT The spread of the Central and South American fern Pityrogramma calomelanos (l.) Link., and its var. aureoflava (Hook.) Weath. ex Bailey, through the Old World is traced. lt appears that both varieties have become naturalised and spread from various centres of horticultural introduction.

INTRODUCTION lt is clear from the relevant I iterature, that the Central. and South American Pityrogramma calomelanos (L.) Link;, and its more temperate var. -aureoflava (Hook.) Weath. ex Bailey, have become naturalised in the Old .World and beyond, the former having become a weed in parts of the Old World tropics. P, calomelanos was introduced to British horticulture from the West lndies in 1790 by Hinton East, Esq. (Aiton 1813) and the var. aureoflava was in cultivation in the Oxford Botanic Garden in 1862 (Hooker 1862). Both var. calomelanos and var. aureoflava, with their respective conspicuous silvery and golden yellow powder on the undersurface of the fronds, have made them both desirable hortiCultural subjects, and also once naturalised, not likely to escape the attention of plant collectors. Unfortunately the taxonomy of this genus was rather confused untii .TryoA's (1962) revision. The probable reasons for the ecological success of the var. calomelanos in West Africa were convincingly presented by Wardlaw (1962). . Although the present author nas been interested in the spread of the var. aureoflava in Natal, South Africa, for manv years· (Aiston & Schelpe 1952) a preliminary study of the spread of both varieties of P. calomelanos through the Old World and beyond became possible during a period of sabbatical leave at the Kew Herbarium during 1973 where most of the data presented here were obtained.

"GEOGRAPHICAL SPREAD The first appearance of P. calome/anos var. calomelanos outside the Americas was in the West African Islands, where it was collected on Principe by Barter in 1858 and on Femando Po by Mann during the period 1859-1863. Its introduction to these islands was probably with tropical American nursery stock brought to these island "plantation colonies" which included coffee, cacao and chinchona (Cortesao 1962). Kuhn (1868) and Hooker and Baker, (1874) only record this taxon from the tropical West African islands outside the Americas and it seems that it only spread :to the West African mainland after the tiJrn of the century. There the earliest record seems to be from Sierra Leone in 1926, but by 1937 Deighton noted on an herbarium label that it was a "garden weed recently established here; it sows itself rapidly". According to available herbarium records it then appeared naturalised in Ghana in 1927, Cameroon Mountain in 1930, Liberia in 1939, Cabinda in 1958 and cSte d'lvoire in 1959. By 1962, it had become a nuisance in Cameroons plantations (Wardlaw, 1962). In East Africa, P. calomelanos was recorded from the Eastern Usambaras in Tanzania in 1914, possibly as an escape from thebotanic gardens at Amani founded in 1903. lt was found at Sese, Uganda, 1922, again possibly an escape from cultivation. 102 FERN GAZETTE : VOLUME 11 PART 2 & 3 (1975)

In 1962, isolated specimens were found in tea plantations on Namuli Mountain, Mozambique (Schelpe 1964 ). Thevar. calomelanos was not recorded from the Seychelles by Baker (1877) but had been introduced by 1902, possibly with tropical nursery stock, followed by the var. aureoflava by 1961. The earliest record of the var. calomelanos in Madagascar found, was collected at Andevorante, Fanovana in 1912, but 20 years later, Christensen ( 1932) noted it as i•very common". Neither varieties were recorded by Baker (1877) for Mauritius or by Cordemoy (1895) for Reunion. Tardieu-Biot (1960 ) regarded P. ca/omelanos as havir1g been introduced to Mauritius and Reunion but dated specimens show that the var. calomelanos was definitely on Mauritius in 1967 and the var. aureoflava on both Mauritius and the Comoros in 1968. The most recent report from Reunion (Sauzier, 1973, pers. comm.) indicates that the var. calomelanos is present and the var. aureof!ava abundant there. As regards South Africa, the Transvaal record of P. calomelanos by Sim ( 1915) is doubtful and cannot be checked as the. specimen was lost in the post. However, there is a definite record of the var. aureoflava, collected in Natal by Gerard in 1865, at Kew, most probably an escape from cultivation. By 1914, this variety had become naturalised in a railway cutting near Durban (Sim 1915) and has since spread to road cuttings as far south as Flagstaff and westwards to the foothills of the Drakensberg. Within the last decade this variety has also apparently escaped from cultivation in the vicinity of Kitwe, Zamb[a (Schelpe 1970). As there is no mention of P. aalomelanos by either Clarke (1 880) or Beddome (1892) it is most probable that it had not become naturalised in India, Ceylon or Malaya at that time. The earliest available record of the var. calome/anos in th is region is from the Terai in the foothills of the Sikkim Himalaya in 1898, to which it could have been introduced as a horticultural subject from the Royal Botanic Gardens in Calcutta. In 1913 it was recorded from the Tista Bridge and by 1933 had reached Kalimpong. Mehra (1932) reported that it had "run wi ld within recent times in Sikkim" and Verma (1966) noted its distribution from the Teesta to Kalimpong and Gangtok. In the meantime it had become naturalised in the states of Bihar and Orissa. As early as 1922, Blatter and 0' Almeida (1922) reported that it had "run wild in and about Bombay" and also reported it from the Nilgiris and Ceylon. lt had apparently become naturalised in the Agri-hort gardens in Madras at the turn of the century and appeared at Tr ivandrum in 1954. Meanwhile the var. aureoflava was reported by Blatter and D'Aimeida (1922, as "var. chrysophylla") as "becoming naturalised in the Nilgiris". In the Far East, the first available record of the var. calomelanos is in 1893 from Buitenzorg, Java, where the famous botanic garden was founded in 1817. Its presence there was recorded by Raciborski (1 898) and a decade later Van Alderwerelt van Rosenburgh (1908) recorded both varieties as "much cultivated and escaped from gardens; very common in the neighbourhood of european settlements". Accordi ng to Backer and Posthumu s (1939). the var. calomelanos had become naturalised in the Javanese lowlands up to 1300 m and the var. aureoflava (cited as P. ta rtarea var. ochracea) naturalised from we st to east Java between 800 m and 1600 m. In the meantime the var. calomelanos had established itself by 1906 on the remnants of the island of Krakatoa, which exploded in 1883, almost certainly by wind dispersal of spores. By 1928, it had spread to northern Sumatra and to Bawean I siand about 50 miles north of Java. An almost certain case of long distance wind dispersal of spores of the var. calomelanos in this region was its appearance as a single sporeling on Christmas Island in 1968, some 200 miles south of Java, in a locality inundated by salt spray for three months, and in which no other ferns were present. SCHELPE: 1HESPRE AD OF PITYROGRAMM A CALOMELANDS 103

In Malaya, the var. calomelanos appeared in a coffee plantation in Perak in 1900 and was later collected in Negri Sembilan in 1906, in Selangor in 1914, in Singapore in 1920, in Paharig in 1922, and in Kelantan in 1935. lt apparently spread further north into southern Thailand about 1923 and reached north-western Thailand by 1958. In th.e meantime it was recorded from Cambodia in 1924, Hongkong in 1926, and Hainan in 1927. lt was collected in Sabah in 1895 and had become a weed in Manila Hemp plantations there by 1962 (Wardlaw 1962), as well as appearing in Sarawak in 1922. After its apparent first record in the Philippines in Luzon in 191 6, it seems to have spread to south-western Mindanao by 1961. Eastwards from Java, the var. ca/omelanos was collected on Halmahera in 1951 and on Ceram in 1970, in the Moluccas. Further east, this variety apparently became naturalised in gardens at Wau, New Guinea, in 1963 and spread to logging roads by 1970. Another apparent example of long distance dispersal was the establishment of this variety in 1953 on lava flows on Mount Lamington, Papua, two years after its eruption. The latest phases of the spread of this taxon are its appearances at Rabaul, New Brita in, in 1962 and in the Whitefield Range of northern Queensland, Austral ia, in 1964. In conclusion it seems that typical P. calomelanos and its var. aureoflava have been introduced to new regions either accidentally together with tropical economic plants or intentionally as horticultural subjects. Their naturalisation and subsequent spread from such foci of introduction has clearly been encouraged by the establishment of plantations and by roadworks, the var. calomelanos spreading mainly in the wet tropics and the var. aureoflava in more tem perate to subtropical environments.

REFERENCES

AITON, W. T. 1813. Hortus Kewensis ed. 2, 5. London. . ALSTON, A.H.G. & SCHELPE, E.A.C.L.E. 1952. Anonnotated check-list of the Pteridophyta of · Southern Africa. J. S. A fr. Bot. 18: 153- 176. BACKER, C.A. & POSTHUMUS, 0. 1939. Varenflora voor Java. Buitenzorg. BAKER, J.G. 1877. Flora of Mauritius and the Seychelles. London. BEDDOME, R.H. 1892. Supplement to the Ferns of British India, Ceylon and the Mala y Peninsula. Calcutta. BLATTER, E. & D'ALMEIDA, J.F. 1922 The Ferns of Bombay. Bombay. CHRISTENSEN, C. 1932. Pteridophyta of Madagascar. Dansk. Bot. Arkiv 7: 113. CLARKE, C.B. 1880. A review of the Ferns of Northern India. Trans. Linn. Soc. Lond. ser. 2, 1: 425-61 1. CORDEMOY, E.J. de. 1895. Flore de i'lsle de la Reunion. Paris. CORTEsAO, A. 1962. The Portuguese discovery and exploration of Africa. Comptes Rendus tve Reunion A.E. T.F.A. T. (Lisbon): 21 -40. HOOKER, W.J. 1862. Garden Ferns. London. . HOOKER, W.J. & BAKER, J.G. 1874. Synopsis Filicum Ed. 2. London. KU HN, M. 1868. Filices Africanae. Lipsiae. · MEHRA, P.N. 1932. Ceropteris calomelanos in Sikkim. J. lndian Bot.Soc. 11: 340--'-341. RAC I BORSKI, M. 1898. Die Pteridophyten der Flora von Buitenzorg. Leiden. SC HELPE, E.A.C.L.E. 1964. Pteridophyta collected on an expedition to Northern Mozambique. J. S.Afr. Bot. 30: 177-200. SCH ELPE, E.A.C.L.E. 1970. Flora Zamb esiaca: Pteridophyta. London. SI M, T.R. 1915. The Fems of South Africa. E d.2. Cambridge. TA RDIEU-BLOT, M.L. 1960. Les foug�res des Mascareignes et des Seychelles. Not. Syst. 16: 151-201. TRYON, R. 1962. Taxonomic Fern notes 11. Pityrogramma including Trism eria and Anogramma. Contrib. Gray. Herb. 189: 52-76. VAN ALDERWERELT VAN ROSENBURGH, C.R.W.K. 1908. Malayan Ferns. Batavia. 104 FERN GAZETTE: VOLUME 11 PART 2 & 3 (1975)

VERMA, S.C. 1966. A note on Pityrogramma calomelanos in India. Buii. BotSurvey India 8: 99- 1 00. WARDLAW, C.W. 1962. A note on Pityrogramma calomelanos (l.) Link, a fern nuisance in Cameroons plantations. J. Ecol. 50: 129-131. FERN GAZ. 11(2 & 3) 1975 105

/ A PHYTOGEOGRAPHIC ANALYSIS OF CHOCO PTERIDOPHYTES

DAVID B. LELLINGER U.S. National Herbarium, Smithsonian Institution, Washington, DC 20560, U.S.A.

ABSTRACT The north western Pacific portion of Colombia (the Depto. del Choco) is of crucial phytogeographic importance because it lies at the juncture of North and South All)erica. The Choci) floristic region is bounded by the Pacific Ocean on the west and by drier areas and by the Andes (at about 1000 m altitude) in the east. To the north it intergrades with the lowland and lower montane areas of Panama and Costa Rica. To the south it intergrades with the lowland and lower montane areas of southern Colombia and Ecuador. Its boundaries can be set somewhat arbitrarily on the north at the Panama border and on the south at the Serranfa Los Chancos north of Buenaventura. Considered as a whole, the pteridophytes of the Choco ftoristic reg ion are most like those of Costa Rica and Panama and secondarily like those of sou thern Pacific coastal Colombia and Ecuador.

INTR ODUCTION The Choco region of north western Colombia is one of the biologically least known areas in the New World tropics. Botanical collecting has been hampered by the seasonally or continually wet climate, the permanently inundated and seasol)ally flooded areas, the scarcity of roads, and the absence of dependable means of transportation (Lellinger & de la Sota 1972). The region is important for distributional studies of both plants and animals, but data sufficient for such studies'exists only for mammals and birds (Haffer 1970) and for pteridophytes. The major features of the region (Fig. 1) are the river basins (Atrato, Baudo, San Juan, and Leon, the last of which drains part of the Depto. de Antioquia into the Golfo de Urbaba) and the mountain ranges (Cordillera Occidental of the Andes and its outlier Lorna del Cuchillo, the Pacific coastal Serranfa de BaudO, the Serranla del Darien, the Cerros de Quia, and the €ordillera de Jurado, the last three of which lie along the Panama border). Daytime temperatures are high throughout the Choco, although the cloud cover moderates them somewhat. From Bahla Solano and Riosucio northward a monsoon climate with a December to March dry season prevails because the northern tradewind belt shifts to the south and sends dry air south from the Golfo de Uraba towards the Pacific coast (Haffer 1970: 611 ). The remainder of the region is more or less constantly wet, with the most rainy area (8000-11000 mm annually) between the Rio San Juan and Rio Atrat

maritima Hieron., Gra mmitis rhizophorae (Copel.) Morton, Lycopodium trianae Hieron., Pteris daguensis (H ieron.) Lellinger3, Selagine/la longissima Baker, S. seemannii Baker, S. tomentosa Spring, Te ctada acutiloba (Hieron. ) Maxon, Th elyp teris in signis (Mett. ) Ching, Tr ichomanes daguense Weath., and seven new species. Nine other species extend southward into Ecaador .(Cnemidaria ewanii (Aiston ) Tryon, Ctenitis biserialis (Baker) Lellinger4. Elaphog/ossum g/ossophy/lum Hieron., E. setigerum (Sodiro) Diels, Grammitis a/sopteris Morton, Polybotrya lechleriana Mett., Te ctaria riva/is (Mett. ) C. Chr., Th e/yp teris elegantula (Sodiro) Alston, and Trichopteris pha/aenolepis (C. Chr.) Tryon. The distribution patterns cited above show that in species terminating at least one end of their range in the Choci5"floris tic region, 29% are Choco endemics, 17% are distributed to the north, and 54% are distributed to the south. The north-to-south ratio is about 1:3. There is no marked difference in north or south distribution between coastal plain or lowland species and upland species. The boundary of the Chad) floristic region on the north can be set at the Panama border because of the fewer species ranging north, compared to those ranging south. In addition, the Pcia. del Darien and areas to the west are drier than most of the Depto. del ChocO. Many Choco species that are also found in Costa Rica are confined to the relatively wet areas of Panama, usually in the Atlantic side of the low mountains. Although more of the aforementioned species range to the south, Cuatrecases (pers. comm. ) believes the limit of the Choc6 floristic region on the south is the Serranla Los Chancos above the Rio Calima in the Depto. del Valle north of Buenaventura. lt is likely that the flowering plants, which tend to have narrower distributions than the pteridophytes, demonstrate the boundary better. Also, the rainfall pattern to the south is more even throughout the year, and the vegetation correspondingly more diverse.

INTRA-CH OCO PHYTOGEOGRAPHICAL RELA Tl ONS HIPS The Chocb pteridophyte flora is not entirely uniform, which is to be expected in an area some 475 km long. In order to understand the differences in the pteridophyte flora within the Depto. del Choc6", the distribution of species found in several scattered areas lying at various altitudes within the Depto. del Choco was studied. For purposes of analyzing the variation, all 'specimens studied were grouped into 16 areas, 11 from the lowlands below 1000 m altitude or from the Serranla de Baudo and five from the Andean slopes of the Choco lying above 1000 m altitude. Figure 1 shows the 11 lowland areas. The five upland areas are from nar San Jose del Palmar and La Mansa in the Cordillera Occidental. The name of each pteridophyte taxon (species, subspecies, variety, of form ) was

3 Pteris daguensis (Hieron.) Lellinger, comb. nov. BASIONYM: Pteris orizabae var. daguensis Hieron. Bot. Jahrb. Engler 34: 495. 1904. TYPE : Banks of the R (o Dagua, Depto. del Val le, Colombia, alt 20G-500 m, Lehmann 8933 (B not see n; isotype US ). This species is close to P. orizabae Mart. & Gal. in frond form, but differs in being less coriaceous, in having more pointed segments, and in lacking the indurated marginal teeth at the apex of each vein. The laminae are ea. 5G-80 cm long and about the same distance wide. The basal pinnae are bipinnate or trip innate, but the superior pinnae are only pinnatifid. The pinnae and pinnules are petiou late (5-10 mm), linear-lanceolate, cu neate or narrowly decurrent at the base, and caudate at the apex. This species is known to me only from the type, from the Santa Marta mountains, from Loma del Cuch illo in the Depto. Choco, and from along the R (o Cauca near Puerto de Ios Pobres in the Depto. de Antioquia. 4 Cten itis biserialis (Baker) Lell inger, comb. nov. BASIONYM : Polypodium biseria/e Baker, Syn. Fil 309. 1867. TYPE: "A '"!des of Peru and Ecuador", Sp ruce (K not see n). , r m G) r c r :::0 m z Cl m !'l ll (') 2 3 4 5 6 7 8 9 10 11 13 14 0 12 15 16 area number and name )> ... , ;;: 360 324 24 7 153 216 181 200 073 055 082 021 020 034 016 030 1 iD northern Choco ::t: e. -< 0 162 158 287 170 250 152 145 119 017 043 052 035 039 2 -i ::I upper Rio San Juan 0 Cl 303 198 244 213 087 059 063 035 05 7 066 017 007 3 interior Bahia Solano & vie. m li-g 0 !2. � e Cl 3 1;" 098 113 143 068 090 062 059 053 038 012 010 4 Serrania del Darien foothills m-· ll _ .. 0 )> 0 , "'C ::I... 121 153 046 061 015 020 037 026 000 010 5 Loma del Cuchillo 0 0 ...... ::t: :r 3 V r+ m 176 120 081 09 1 021 000 057 026 022 6 coastal Bahia Solano & vie. n "'-· ...... '0 )> c)(" 117 061 032 020 058 071 012 022 7 Rio Mutata () z 0 )> 0 .c �. u r � ..... 097 056 000 047 048 046 051 8 Alto del Buey montane forest -< �a; Ul CT 040 051 030 194 lOO 100 9 Al to del Buey mo ssy forest Ul .....CD m ... 0 0 .. ., ... .. intra-Choco ave. r = 050 095 045 060 050 10 Lower Quibdo--Bolivar Road .. "' n ... -· ::t: =t' ::::l 024 070 031 000 11 Lower Rio San Juan & vicinity CD 0 ::I .... n = = c:: ..=r Choco--Andes ave. r 050 12 San Jose del Palmar cemetery o, 3 0 , CTC 13 0 -i .... Middle Bolivar--Quibdo Road 0 m '"'"C .,... ll I" m 1\ ... intra-Andes ·ave . r = 034 14 San Jose del Palmar ridge .... 0 m U) 3 QJ 0 .. 15 La Hansa ridge "" , ::I ::t: .... � -< 0 16 Serrania de Lo s Paraguas -i c. m !!. Ul (') =r 0 n P\ -1 ... =r 0 .. (I) 110 FERN GAZETTE: VOLUME 11 PART 2 & 3 (1975) written on an edge-punched card and the 16 areas each assigned a position along the margin. These were notched for each area in wh ich the taxon wa� known to occur, either as a herbariu m specimen or as a sight record made by Lellinger and de la Sota. The presence data were then compared for each pair of areas, using a familiar coefficient of association: (++)/(+-)+(-+), which is the number of taxa present in both areas divided by the number found in one or the other of the areas, but not both areas. With respect to the limits of the Chocb floristic region in the Andes, the average of all the correlation coefficients linking the 11 loc'alities lying below 1000 m altitude or in the Serranfa de Baud6 (Fig. 2) was r=0.156. The average of all the correlation coefficients linking the five localities lying above 1000 m altitude on the Andean slopes of the Departamento del Chocb was r=0.060. But the average of all the correlation coefficients linking the former with the latter was only r-0.040. This indicates that the lowland (Choc6) pteFidophyte flora is distinct from the upland (Andean ) pteridophyte flora. The matrix of coefficients of association for the 11 non-Andean areas (Fig. 2) was used to produce a phenogram (Fig. 3) showing the relative similarity of the areas. An unwe ighted group method was used to ca lculate the values joining the branches of the phenogram, starting with the highest mutual coefficient (r=0.440) and working downward. The value of the coefficient joining a new area to the group of areas already placed on the diagram was the average of the coefficients between all pairs of members of the group, including the new one. An examination of the phenogram shows some interesting relationships among the various areas. The R(o Mutata hillside and floodplain area (area 7) and the montane forest on the slopes at the base of Alto del Buey (area 8) are close geographically, and all are undisturbed, primary forest. The mossy forest on the upper slopes towards the summit of Alto del Buey correlates most closely with the lower Andean slopes at 750 m altitude (area 10), more so than with the pteridophytes of the montane forest slopes at the base of Alto del Buey (area 8). Thus, the pteridophyte flora near the summit of Alto del Buey definitely belongs to the Chad) floristic rl:!gion, even though it is above the 1000 m Choc6-Andean pteridophyte flora boundary. There are, however, some Andean species, like 8/echnum occidentale L. and Lophosoria quadripinnata (Gmel.) C. Chr., that set it apart from lower altitudes in the Choco. The Serranfa de Ba�db, being surrounded by lowlands, is probably warmer for its altitude than are the Andes, which rise much higher. The altitude of Alto del Buey is given on most maps as 1810 m, as Lellinger and de la Sota's specimen labels indicate that the mossy forest is at 1450-1 810 m, but Haffer (1970, p. 608) suggests that 1342 m may be more nearly correct. The lower R (o San Juan (area 11). which is the southernmost study area in the Choco, is le :::l "tl 0 J: � ., 0.267 4 Serran1a del Darien foothills < 3 ....---- -1 0 -0 Gl - m - 0.246 5 Loma del Cuchillo 0 Gl ., r---- Jl CD... ., l> en "tl :i" J: .... 0.222 0 :::r 6 coastal Bah1a Solano & vi e. CD l> z 0 CD l> 'C r ., < 0.440 en iit 7 R1o Mutata 3 r 1ii "' l 8 Al to del Buey montane fores t 0 :::l .., g ....---- 0 c. J: !!!. 0 (") 0 :::r 0\ g "tl 0\ 9 Alto del Buey mossy forest -1 0.263 m < 0. Jl :i" ICI 0 0 r:T 10 Lower Quibdo--Bol1var Road "tl "' J: 0 < :E -1 - m 8 en 0 3 11 Lower R1o San Juan ., ;::+ ;:;: c 0. riso 0. Joo o.hso c. !" 112 FERN GAZE TIE: VOLUME 11 PART 2 & 3 (1975) the several northern Choc� local ities (area 1) to interior Bahfa Solano near the Pacific coast (area 3) and the upper R(o San Juan (area 2), form a fairly coherent group, with more Panaman ian-Costa Rican species present in the Serran(adel Darie'n foothills than elsewhere in central and northern Choc6.

/ EXTRA-CHOCO PHYTOGEOGRAPH ICAL RELATIONSHIPS Evidence concerning the phytogeographic relationships of the lowland Choco pteridophytes with the pteridophytes of other areas of tropical America is substantial, but not completely conclusive. Almost all the data on extra-Choco distributions were taken from the U.S. National Herbarium collections, with some additions from the literature for the species of a few genera. The data are subject to collecting bias and representation in the herbarium, wh ich probably accounts for the low percentage of Choco pteridophytes recorded from El Salvador and, perhaps to a slighter extent, from Ecuador and Nicaragua. To study the affinities of the Choc6 pteridophyte flora, the New World was divided into 27 areas, each of which is a country or (in the Antilles) an island or a group of islands. Each Chod) taxon card was notched for each area in which it was found. Figure 4 shows the results rounded off to the nearest percent. There is a decided drop-off in the percentage of affinity from Costa Rica to Nicaragua, from Guatemala to Mexico, from Venezuela to the Guianas, from Colombia to Brazil, and from Peru to southern South America. The highest percentages of pteridophyte taxa corresponding between the Choc6 and the other areas are· found in Colombia, Costa Rica, Panama, and Peru. Although no numerical data have been collected, I estimate that about 85% of the Choc6 pteridophyta are known in Colombia from departments other than the Choco. Some extend only into parts of the departments of Antioquia and El Valle that lie within the Choco floristic region, and so are not truly found in non-Choco Colombia. Those that do extend to other parts of Colombia are mostly widespread species that also range beyond Costa Rica and Ecuador. Judging by the north or south distribution of near-endemic Choco species discussed above, the affinities are with the Pacific Coastal phytogeographic areas, which are both adjacent to the Choco geographically and also are more like it in climate. The average of the percentages found in Colombia (estimated at 85%), Ecuador (54%). and Peru (60%) is 66%. The 79% for Costa Rica and Panama is about 20% higher. The affinity of the Choc6 pteridophyte flora appears _to be primarily with that of Costa Rica and Panama and secondarily with that of the Pacific coastal and submontane slopes from Colomb ia to Ecuador.

ACKNOWLEDGEMENTS Most of the data for th is paper came from specimens collected by the Smithsonian lnstitution-Museo de La Plata Botanical Exploration of the ChocO, which was made possible by grants from the National Geographic Society and the Smithsonian Research Foundation, plus financial support of the Comision de lnvestigacion Cient(fica de la Provincia de Buenos Aires. The exploration was aided immeasurably by the staffs of the Institute de Ciencias Naturales at Bogota, the Departmento de Biolog(a of the Universidad de Antioquia at Medellln, and the Institute de Desarrollo de Ios Recursos Naturales Renovables (INDERENA) at Bogota, Medellln, and Turbo. I wish to thank Drs. Jose Cuatrecasas, Helen Kennedy, and Harold Robinson for their valuable suggestions concerning this paper. LELLINGER: A PHYTOGEOGRAPHIC ANALYSIS OF CHOC O PTERIDOPHYTES 113

FIGURE 4. Map showing percentage of Choc6 pteridophytes also known from other countries or Antillean island groups of the New World. 114 FERN GAZETTE : VOLUME 11 PART 2 & 3 (1975)

REFERENCES HAFFER, J. 1970. Geologic cl imatic history and zoogeographic significance of the Uraba region in north western Colombia. Caldasia 10: 603-636. LELLI NGER, D.B. & de la SOTA, E.R. 1972. Collecting ferns in the Choco. Amer. Fern J. 62: 1-8. SOTA, E.R. de la 1972. Las pteridiofitas y er epifitismo en el Departamento del ChocO'(Colombia). Anal. Soc. Cient. Argen tina 194: 245-278. FERN GAZ. 11(2 & 3) 1975 115

STUDIES IN THE SYSTEMATICS OF Fl LMY FERNS: I. A NOTE ON THE IDENTITY OF MICROTRICHOMANES

K. IWATSUKI Department of Botany, Faculty of Science, Kyoto University, Kyoto, 606, Japan.

ABSTRACT

The features of the filmy fern genus Microtrichomanes are described. All the species ' referred to this gen�s are compared with each other from the viewpoint of alliance. Subdivisio.n of the genus is proposed, and the �elationship of each species is discussed, giving critical comments on the current conception that Microtrichomanes stands between Trichomanes s.lat. and Hymenophyllum s.lat.

INTRODUCTION There are two alternative systems in the Hymenophyllaceae currently adopted by pteridotogists: one is the classical two genera system, recognizing Trichomanes and Hymenophyllum in a broader sense, or sometimes separating a few distinct genera in addition to the above two (e.g. Morton 1968); the other is the system proposed by Copeland (1938 1947 ) distinguishing 34 genera. We have still insufficient information about the filmy ferns to conclude which system is natural and correct, and it is intended in this and following papers to elucidate the natural relationship among the species of the Hymenophyllaceae. Microtrichomanes was first recognized as a group by Mettenius (1864) enumerating Tr ichomanes digita tum, T. palma tifidum and · T. dichotomum. SJbsequent authors added further species, increasing the number to ·14 as listed by Morton (1968). Concerning the systematic position of this genus among the filmy ferns, there is general agreement that it is intermediate between Tr ichomanes s.lat. and Hymenophyllum s. lat., although Christensen (1906) included Microtrichomanes in Trichomanes sect. Gonocormus. This intermediate position of Microtrichomanes forms one of the reasons why Copeland could not maintain the above two genera in the broader sense. In his opinion (Copeland 1938: 36) the species of Micro trich omanes "present sufficient evidence of affinity to Sphaerocionium . .. so far as the marginal setae are present". Similarly, Holttum (1955: 87) discussed the hairs and sari of

Micro trichomanes and states that "there· is little doubt th ey are more nearly related to c Hymenophyllum ". After detailed observation on most of the species belonging to Microtrichomanes, I am now inclined to have the same opinion· as Morton (1968 : 199) who says that "it is decidedly heterogeneous and would seem to be polyphyletic", though I have no evidence at hand to support his view that the genus "may have arisen anciently from hybridization between various species of Trichomanes and 1-J.;menophy(lum, perhaps belonging to quite different groups". This is a preliminary report, for the true relationship will be elucidated only after detailed studies on the other "genera". Therefore, no new names are proposed here, and the specific names are used enumerating the species of Microtrich omanes under Tr ichomanes except in the case wh�re no formal combination has ever been given in that way. The names of subdivisions of the filmy ferns are adopted in accordance with the Copeland's system, unless otherwise stated. These subdivisions are o�ten referred to as genera, though they are not used here in the form of binary names. 116 FERN GAZETTE : VOLUME 11 PART2&3 (1975)

t-i

a

d \\h FIGURE 1: Pinnation in fronds in Microtrichomanes, all x 1/2. a, T. dich otomum (Sumatra, lwatsuki et al. S- 1127); b-f, T. flabe/latum b, Sumatra, lwatsuki et al. S-1 163 bis; c, New Gu inea, Kalkmann 5007; d, Borneo, Hotta 3854; e, Thailand, Tagawa et al. T -4823; f, Malaya, lwatsuki et al. M-13234) ; g, T. taeniatum (Tahiti, Grant 3561); h, T. nitidulum (Malaya, lwatsuki et al. M- 13514); i, T. /yallii var. neocaledonicum (New Caledonia, Franc 828).

OBSERVATIONS ON TAXONOMIC FEATURES The following taxonomic characters used to evaluate Microtrichomanes are described and discussed from the morphological and syst�matic point of view.

Rhizome Most species of Micro trichomanes grow on tree trunks and boles and have a slender, more or less wiry rhizome less than 0.2 mm in diameter, brown, and sparsely bearing brown hairs usually about 1 mm in length. In T. francii the rhizome is wiry, about 0.3 mm in diameter, and nearly glabrous except in apical portion where brown hairs are found. In T. vftiense and T. aswijkii, the rh izome is· not wiry, less than 0.2 mm in diameter, and quite densely covered wi th the blackish coarse hairs. The last type of rhizome is found also in Microgonium, while the first type is most frequently known in the filmy ferns especially common in Hymenophy!lum s.lat. IWATSUKI: STUDIES IN THE SYSTEMATICS OF Fl LM Y FERNS 117

a

FIGURE 2: Fronds of Microtrichomanes: a, T. fra ncii (New Caledonia, Franc s.n.), x 2; b, T. vitiense (Fiji, Degener 14607), x 2; c, T. parvulum (Madagascar, Forsyth-Major 203), x 2.

Pin nation The pinnation in fronds is always described as dichotomous for Microtrichomanes, though this is not so in real ity. Figs. 1 and 2 illustrate rather schematically the pinnation pattern of various forms referred to this genus. lt is evident that T. dicho tomum has pinnately divided fronds, and this is quite true for T. fla bella tum, T. ta eniatum and T. lvallii var. neoca/edonicum, though their fronds are short and seemingly digitate. In T. pa/ma tifidum and T. nitidulum the length of fronds is reduced, and the rachis is often very short usually giving the appearance of a dichotomously branched frond. lt may be concluded that the apparently dichotomous fronds have resulted from the extreme shortening of the rachis by the reduction of the length of fronds. This does not necessarily mean that the larger species are primitive; as noted below, T. dichotomum is a specially elongate form of T. flabella tum, and can not be considered as the prototype of the latter. The same kind of variation in pinnation is seen in the various forms of Gonocormus. The pinnately divided fronds have a definite rachis even when the length of fronds is extremely reduced, and typical dichotomy cannot be found even when the number of segments is extremely reduced. Contrary to these species, the fronds of glabrous species are .simple to biforked or dichotomously branched when the number of segments is more .than four. A similar pattern of branching in fronds is fou nd in reduced species of Microgonium, and the simple frond may be referred to the most reduced form in th is series. 118 FERN GAZETTE: VOLUME 11 PART 2 & 3 (1975)

Hairs There are two types of hairs in Microtrichomanes. Setose marginal hairs are found on u�timate segments in T. digita tum, T. flabellatum, T. dichotomum, T. ta eniatum and T. lyallii var. neocaledonicum. They are simple, short and rather thick, usually 0.2 mm long and up to 0.03 mm in diameter, and dark brown to blackish in colour. Hairs of th is type are not found elsewhere in the filmy ferns except in the simple form of the marginal hairs of Didymog/ossum. The other type of hair is that found in T. palmatifidum and T. ridleyi. In T. pa/ma tifidum they are found rather densely on the margin of segments and in some plants also on the veins. T. nitidulum is usually glabrous but occasionally bears the hairs of this type on the margin of segments. Hairs of this type are simple or forked, more or less downy, about 0.5 mm long, and pale brown in colour, and are similar to those found in Sp haerocionium. Various types of hairs in Sp haerocionium are summarized by Morton (1947) with diagrammatical illustration. In all the species of Sp haerocionium, the hairs are pale brown, long and slender, simple to stellate and no species have only simple hairs, e.g. the basally forked hairs are found mixed with the simple hairs as in Hawaiian H. lanceolatum, or many species have the forked and stellate hairs in add ition to the simple hairs as in H. microcarpum. In T. palmatifidum no stellate hairs are observed, though we have no information to distinguish the hairs of th is species from those of Sp haerocionium. Th ere are species quite glabrous and still referred to Microtrichomanes. They are T. aswijkii, T. vitiense, T. francii and T. parvulum.

Sorus The structure of the involucre is obconic to campanulate or tubular; receptacles are long and usually extruded. In T. nitidulum and T. palmatifidum, the involucre is relatively short and obconic in appearance (fig. 3a), though relatively long and campanulate in T. flabel/atum (fig. 3b) and the other species with setose marginal hairs. The variation in the form of involucre is not correlated with the breadth of the segments in general, except in T. taenia tum in which a tubular involucre is found in plants having narrower and longer segments and a campanulate one in plants with broader segments. The mouth is hardly bilabiate except in T. nitidulum and T. pa/ma tifidum, in which the lips of the mouth are slightly extruded over and above the apex of the segments, though the form of lips is subtruncate to very moderately round. In T. aswijk ii the involucre is tubular and the mouth is distinctly dilated as in the case of Gonocormus and Microgonium. Generally the structure of the involucre is referable to that of Tr ichomanes s.lat. lt is interesting in this connection that H. borneense Hook. is described under Hymenophyllum, though this is identical with T. palma tifidum.

COMMENTS ON THE SPECIES Morton (1968) referred 14 species to Trich omanes sect. F/abel/ata, an equivalent of Microtrichomanes. All the species referred to Microtrichomanes are treated even when the same author separated the species from that genus in a later paper. A complete

revisional treatment will be given in a future publication acc· ording to the genera defined there.

1. Trichomanes lyallii (Hook. f. ) Hook. This species was included in Microtrichomanes by Copeland in his 1933 paper but IWATSUKI: STUDIES IN THE SYSTEMATICS OF FILMY FERNS 119

d

Fl GURE 3: Sori and hairs of Microtrichomanes: a, sorus of T. palmatifidum (Sumatra, lwat&Jki et al. S-1163), x 20; b, sorus of T. flabellatum (Malaya, Togashl s.n.). x 15; c, hairs of T. palmatifidum (Sumatra, lwatsukl et al. S-1 163), x 30; d, hairs of T. flabella tum (Thailand, lwatsuki at al. T-83691, x 30. 120 FERN GAZETTE: VOLUME 11 PART 2 & 3 (1975)

transferred to Sp haerocionium by himself in 1938. Morton (1968) placed this in his Trichomanes sect. Flabellata probably following Cope land ( 1933) without detailed study of the plants. This was placed in Microtrichomanes by Copeland on account of the reduced flabellate frond, though the marginal hairs are stellate, or d-type in Morton's scheme of 1947. Tindale (1963) made detailed observation on the size and form of fronds as we ll as the marginal hairs of th is species. lt is evident that this species belongs to Sphaerocionium indicated by the stellate marginal hairs and included cylindrical receptacles.

2. Trichomanes palmatifidum C. M� ller This species has always been treated as an ally of T. digita tum since Mettenius (1864) included this in his group Microtrichomanes, but the hairs on fronds of T. palma tifidum belonging to the types a and b in Morton's scheme are quite different from the marginal setose hairs of T. digitatum group and the same as the hairs found in H. lanceo/atum and also in H. microcarpum mixed with types c and d by Morton. The fronds of th is species appear digitate and are commonly described as such, though this is pinnately compound wi th long segments and a short rachis. This frond construction is also found in T. nitidulum, and relates these species to T. digitatum. There is no example of reduced flabellate fronds in Sp haerocionium unless T. /ya/lii is included in that genus. The flabellate or digitate appearance of the fronds of T. pa/ma tifidum is interpreted as the result of the reduction of length from more slender, pinnately compound fronds. The involucre is obconic with somewhat bilabiate mouth (fig. 3a ), and is somewhat similar to that of T. f/abel/atum with a campanulate to tubular involucre with a slightly dilated mouth. The extruded receptacle is distinct from that of Sphaerocionium. The morphology of the receptacles will be discussed in detail in a separate paper and here it is noted only that the receptacles of T. palmatifidum and T. nitidu lum are not too long, and that receptacles can be extruded in various species of Meringium. T. pa/ma tifidum can be referred to as an ally .of Sp haerocionium in spite of having the extruded receptacles.

3. Trichomanes nitidulum van den Bosch This species is usually described as glabrous, though thedetai led observation of ample material shows that long downy hairs are sometimes found at the margin of ultimate segments. The morphology of these hairs is quite identical to those of T. palma tifidum. Except for this "glabrous" nature, T. nitidulum is the same as T. pa/ma tifidum, and as the density as we ll as the occurrence of hairs are rather variable in that species, T. nitidulum seems to be better treated as a variety of T. palmatifidum.

4. Trichomanes ridleyi Copeland This is distinguished from T. palma tifidum by long stipes and rather irregularly branched fronds with sparse hairs, and is known only from the type collection from G. Tahan in Malaya, where T. palma tifidum has repeatedly been collected. Some of the latter plants are similar to T. ridleyi (e.g. Wray & Robinson 5401 in SING). Holttum (1955) considers it "possibly a hybrid" between T. digitatu/]1 and T. palmatifidum, though by comparing the variation of pinnation pattern of T. palmatifidum and T. nitifulum it is suggested here that T. ridleyi is only a variant of the latter. IWATSUKI: STUDIES IN THE SYSTEMATICS OF FILMY FERNS 121

5. Trichomanes digitatum Sw. T. digitatum in the most strict sense, excluding is T. flabel/atum of Southeast Asia, is restricted to East African Islands, in Mauritius and Madagascar. The African form differs from T. flabellatum in its smaller size and the more simple aspect of fronds which have 1-5 segments digitate in appearance. In T. f/abel/atum, mature fronds with several segments are occasionally observed, although they are usually found on the same rhizome with the larger fronds with many segments.

6. Trichomanes flabellatum van den Bosch Although I have insufficient data about the plants in East African Islands, I would agree partly with Christensen (1932) in distinguishing the African form from the East Asiatic form (including the typeof T. flabella tum ). The difference is found only in the degree of development of the fronds, and they might be better segregated in the rank of sub-species considering the difference in their distribution. Thus confined, T. flabe/la tum is recorded only in South-east Asia: in Sumatra, Peninsular Thailand, Malaya, Borneo, Philippines, Java and New Guinea. I have not seen the type of T. piliferum v.A.v.R., and fol low here Backer and Posthumus in reducing it to T. flabellatum.

7. Trichornanes dichotomu m Kunze In its largest form, th is species is more than 20 cm in length and the digitate construction of fronds becomes elongate. As such it is distinct from T. flabellatum, though there are intermediate forms between the two extremes. lt may be speculated that these long fronds are evolved from short fronds of T. f/abella tum, and with the elongation of fronds the ultimate segments become longer with the tuijular involucres bearing long extruded receptacles. This is not distinct specifically from T. flabellatum and is better treated as a variety of the latter.

8. Trichomanes taeniatum Copeland This is similar to and possibly should be regarded as a subspecies of T. flabel!atum which differs from it in usually having more numerous segments which are slender, green or rarely dark in dried specimens, with blackish and distinct veins. When the ultimate segments are slender, the involucre becomes longer and tubular in appearance. Copeland (1933) noted the·variation of th is species; the Samoan plants referred by him to T. digitatum belong here, and its range includes the Society, Solomon and Marquisas Islands.

9. Trichomanes lyallii var. neocaledonicum C.Chr. There are two taxa of Microtrichomf!nes represented in New Caledonia. One was described by Christ as T. francii (see below). The other belongs to a group of T. digitatum both in pinnation and setiferous margin of segments. lt is similar to T. flabel!atum but differs in having an obconic involucre and the smaller fronds usually of equal length and breadth. The hairs at the segment margins are rather slender and not so dark as typical T. flabellatum. The figure given by Brownlie (1969: pi. 12, fig. 6) is this setiferous form, not T. fra ncii as stated. T. /yal/ii sensu Fournier (1 873), T. /yallii var. neocaledon'icum C.Chr. and T. digitatum sensu Compton (1922) are names given to this latter form, though I am not sure at present whether T. cuneatum Christ is also this taxon or not. Cpmpton referred Rosenstock Exsicc. n. 51 to T. cuneatum, and that I have examined (in herb UC) is this taxon, but I have not so far seen Christ's type 122 FERN GAZETTE : VOLUME 11 PART 2 & 3 ( 1975) and his incomplete description is insufficient for determination. Anyway, this setiferous form is probably best treated as a variety or subspecies of T. digitatum and var. neoca/edonicum is available as the correct name at varietal rank.

10. Trichomanes francii Christ This is a completely glabrous form: rhizome wi ry, nearly glabrous in older ones; segments 1 to 8, usually shorter than the breadth, each fork with wide sinus; involucre ' obconic, the mouth hardly dilated and with subtruncate apex. lt is so far endemic to New Caledonia.

11. Microtrichomanes zamboanganum Copeland I have not seen any material of this species, and place it at present close to T. francii only from the description and notes given by Copeland (1952, 1958).

12. Trichomanes parvulum Prior I have not examined the type material of various names referred to this species, and so discuss here the affinities of this taxon but not the nomenclature. The dichotomous branching type of frond is found in this species and in T. franc ii (cf fig. 2a,c). though it is sometimes deformed in the latter, especially in the larger forms. The frond pattern is simi lar to Gonocormus but the latter differs in its sorus and proliferous habit. The involucre of Gonocormus is usually campanul ate with a dilated mouth, whereas that of T. parvulum is obconic with a bi labiate mouth bearing round and irregularly dentate lips.

13. Trichomanes vitiense Baker lindale (1963) gave a detailed description and notes on this species, and I have little to add. lt differs decidedly from T. digita tum s.lat. in completely lacking hairs on the fronds. This is similar to T. aswijk ii in having a slender rhizome densely covered with blackish hairs. The thin fronds remain green in the dried condition or eventually become pale brown. In these features as well as in general habit, this resembles species of Microgonium, and Tindale compared th is actually with T. motleyi. I have no conclusive remark at present to th is affinity, for the false veinlets in the filmy ferns are now under investigation and are still little known from th e phylogenetic point of view. Besides Microgonium, th is and the following species are comparable with T. fra ncii and T. parvulum in some features. In the morphology of the involucre, T. vitiense and T. aswijk ii resemble Gonocormus.

14. Trichomanes aswijkii Racib This is similar to T. vitiense but differs from it in: stipes shorter, less than 1 mm long or sessile; the base of simple frond narrowly cuneate; the segments 1 to 4 (very rarely 6) 2.5-3 mm in breadth; the rhizome slender, bearing dense blackish hairs. Compared with T. francii the segments do not branch so widely and the involucre is tubular with a distinctly dilated mouth. This and the preceding seem to be closely all ied and are compared with Microgonium tentatively; the alliance however, is obscure.

15-16 Trichomanes barklianum Baker and Trichomanes liberiense Copeland Both of these African taxa were included by Copeland in Microtrichomanes in 1933 IWATSUKI: STUDIES IN THE SYSTEMATICS OF FILMY FERNS 123

but referred to Didymoglo$sum· in 1938. In the absence of false .veinlets and presence of the marginal hairs, they seem to be similar to T. digita tum, but I cannot comment · ' further at this stage. · '.

SUMMARY Thespecies usually referred to Microtrichomanes may be grouped as fol lows : 1: Species included in, or compared with, Sphaerocionium : T. lya/lii, T. palmatifidum, T. rid/eyi and T. nitidulum. 2: Micro trichomaries s.str.: T. digita tum, T. flabe/latum, T. dichotomum, T. taeniatum and T. /yallii var. neoca/edonicum. 3: Species with dichotomous fronds comparable with Gonocormus: T. francii, M. zamboanganum (?) and T. parvulum. 4: Species close to 3 or comparable with Microgonium: T. vitiense and T. aswijkii. 5: Species probably belonging to Didymog/ossum : T. barklianum and T. /ib eriense. As summarized here, Microtrichomanes has usually been defined by the dwarfed frond construction and includes a variety of taxa, which may be regarded as belonging to various species groups or genera. lt is quite natural that such a heterogenous "genus" relates both to Trichomanes s.lat. and to Hymenophyllum s.lat. However, if group 1 is removed, Microtrichomanes has no affinity with Hymenophyllum s.lat. and ·belongs wholly to Trichomanes s.lat. even if groups 2-5 are amalgamated into a single unit or genus. The real position of each species group wi 11 be elucidated by further studies, especially on the other groups of the filmy ferns, although I have shown that Copeland had little basis for segregating Microtrichomanes from the classical twoiJenera system of Hymenophyllaceae.

ACKNOWLEDGEMENTS This study is based ch iefly on the collection kept in the herbarium of Kyoto Universit_y (I

REFERENCES BROWN LI E, G. 1969. Hymenophyllaceae in Fl. Nouv.-Caled. Depend. 3: 87-110. CH R ISTENSEN; C. 1906. Index Filicum. Copenhagen. CHRISTENSEN, C. 1932. The pteridophytes of Madagascar. Dansk Bot. Ark. 7: 1-253. COMPTON, R.H. 1922. A systematic account of the plants collected in New Caledoniaand Isle of Pines 11 : Pteridophyta. J. Linn. Soc. Bot. 45: 435-462. COPELAND, E.B. 1933. Trichomanes. Phi/. J. Sci. 51: 119-280. COPELAND, E.B. 1937. Hymenoph ylfum. Phi/. J. Sci. 64: 1-188. COPE LAND, E.B. 1938. Genera Hymenophyllacearum. Phi/. J. Sci. 67: 1 -110. COPELAND, E.B. 1947. Hymenophyllaceae in Genera Filicum: 31-44. Waltham, Mass. COPELAND, E.B. .19 52. New Philippine ferns IX. Phi/. J. Sci. 81: 1-47. COP ELAND, E.B. 1958. Hymenophyllaceae in FernFl. Phi/. 1: 46-82. Manila. FOURNIER, E. 1873. Filices NCJ.tae-Caledoniae. Ann. Sci. Nat. V. 18: 253-360. HOLTTUM, R.E. 1955. Trichomanes in Rev. Fl. Mal. 2:. Fernsof Malaya: 72- 109. Singapore. IWATUSKI, K. 1968. Contributions to the classification of the filmy ferns (2). Acta Ph ytotax. Geobot. 23: 111 125 (in Japanese). 1, METTENIUS, G. 1864. Uber die Hymenophyllaceae. Abh. Sachs. Ges. Wiss. 7. MORTON, C.V. 1947. The American species of Hymenophyllum section Sp haerocionium. Contr. U. S. Nat. Herb. 29! 139-201.- 124 FERN GAZETTE: VOLUME 11 PART 2 & 3 (1975)

MORTON, C.V. 1968. The genera, subgenera and sections of the Hymenophyllaceae. Con tr. U.S. Nat. Herb. 38: 153- 214. SLEDGE, W.A. 1968. The Hymenophyllaceae of Ceylon. J. Linn. Soc. Bot. 60: 289-300. TAR DIEU-BLOT, M.L. 1951 . Hymenophyllac:eae in Fl: Madag. Com. 38 fam .: 1-38. TINDALE, M.D. 1963. Hymenophyllaceae. Contrib. N. S. W. Nat. Her{J., Flora ser. no. 201: 1-49. VAN DEN BOSCH, R.B. 1861. Hymenophyllacearum J81fanlcae. Verh. koninkl. Akad. We tersch. Amsterdam 9: 1-64. FERN GAZ. 11 (2 & 3) 1975 125

A HYBRID POL YP ODY FROM THE NEW WORLD TROP ICS

W.H. WAGNER JR. and FLORENCE S. WAGNER Department of Botany, University of Michigan, Ann Arbor 481 04, U.S.A.

ABSTRACT

Polypodium leucosporum Klotz., from Jamaica is shown to be a hybrid between P. la nceolatum L., ofte n placed in Pleope�tis, and P. thyssanolepis A. Br., a component of Marginaria or Polypodium s. s. The variation, especially of frond dissection and venation, is given and the hypothesis that there are genetically different forms of P. x leucosporum and that other species and/or cytotypes are involved in their origin is discussed.

INTRODUCTION Most hybrid polypodies that have been found are crosses between sectors of the same species or between similar sibling species (e.g. Lang 1971 ). Wide hybrids (crosses between strongly distinctive elements) are very rare in polypodies, as they are in all groups of ferns. Such hybrids have special interest, however, in helping us assess systematic relationships and establish homologies. Th is is particularly true of those hybrids which comb ine characteristics of simple-leaved and divided-leaved taxa which have widely differing patterns of venation and soriation. In the Old World polypodioid genus Co/ysis, Lellinger (1968) discussed and figured C. x shintensis (Hayata) H. Ita, a plant morphologically intermediate between C. wrigh tii (Hook. ) Ching, with simple blades, and C. elliptica (Thunb.) Ching, with pinnate blades. In Polypodium and its segregates in the New World, Weatherby (1935 ) described P. bartlettii as a possible hybrid between the simple-fronded P. poly/epis Roem. (= P. peltatum Cav.) and the pinnatifid-fronded P. polypodioides (L.) Watt. He compared this fern with the South American and West Indian P. leucosporum Klotzsch, which Vareschi (1969 ) later interpreted as the cross between P. /anceolatum L. and P. th yssanolepis A. Br., plants frequently placed in different subgenera or genera. All of these presumed hybrids exemplify what we have termed "hybrid irregularity" (Wagner 1962 ) in which the intermediate plant expresses the parental characteristics asymmetrically. Here we wish to discuss the morphology and cytology of one of those irregular ferns, basing our studies on both field-collected and greenhouse-grown specimens. We were especially interested in studying the intermediate characters and the extent or irregularity, emphasizing frond habit, venation, scales, sari, and chromosomes. Herein the plant involved will be given a new status, based upon the facts presented below; its name should now be as follows : ,,. Polypodium x leucosporum Klotzsch, Linnaea 20: 404 (1847 ),pro sp. P/eopeltis leucospora (Kiotz. ) T. Moore, Index Filicum : 77 (1857 ). Lepicystis leucospora (Kiotz.) Diels, in Engler & Prantl, Die natUrl ichen Pflanzenfamilien 1 (4): 324 (1899).

MATERIALS AND METHODS We encountered a large colony of P. x leucosporum in Surrey Co., St Andrews Parish, Jamaica, about one-third of a mile from the Cinchona Gardens near the trail to New Haven Gap. Specimens were collected on 2 March 1970 (Wagner 70100, MICH), together with associated samples of P. lanceolatum (70101) and P. th yssanolepis (70 102) growing in the same place. We found our specimens on rocks and at bases of 126 FERN GAZETTE: VOLUME 11 PART 2 & 3 (1975) small trees at the edge of a field. The habitat of P. x-leucosporum and similar fo rms everywhere in the New World tropics seems to be terrestrial rather than epiphytic, usually on disturbed rocky ground with more or less shade, in mountainous regions at altitudes of 1300-3000 m. Statements gleaned from herbarium labels include such descriptions as the following: "on a wa ll", "on limestone rocks", "forest, on ' the ground", "dry rock wall", "on trunk of tree on hill", "on top of boulder", "on stump at border of cultivation", and "creeping on moss over boulders". Living materials of all three polypodies were collected for cytological study. lt turned out that P. x leucosporum flourishes under our" greenhouse conditions and produces a vigorous potted plant. In culture it forms leaves over twice as large as those from the wild and variations appeared that are not yet known in herbarium collec;:tions. 'Hybrid fronds were consistently more robust than those of the parents, and some examples are shown in fig. 2 e-f.

MORPHOLOGY

Po/ypodium x /eucosporum has consistently fewer lobes than P. thyssanolepis (fig 3 ). Furthermore, P. x leucosporum has numerous "mini-lobes", i.e. lobes only half or less the size of the adjacent ones. A majority of these miniature lobes are merely shallow concavities in the blade outline. Some herbarium collections (fig la) have only this type of lobing. The graph (fig 3) shows averages of many leaves grown in uniform culture; it shows that P. x leucosporum is intermediate in lobation between P. thyssanolepis represented by the upper line and P. lanceolatum the base line, in the last there being no lobation at all. The major segme nts in P. x leucosporum tend to be longer and more pointed than in P. th yssanolepis. They are broadly adnate at base rather than contracted. They tend to be borne at a steeper angle and the vein departures at th eir bases have a different configuration. Rather than turning abruptly outward as in P. th yssanolepis, they run ·upwards for 4-1 0 mm and only then bow outwards. Related to this is the expansion of the laminar wing along the' rachis, which is considerably wider in the hybrid than in P. thyssanolepis. • The most distinctive feature of frond structure in the hybrid is, of course, the asymmetrical outline. Irregularities may occur anywhere in the lower two-thirds of the blade, below the entire-margined tip, but two zones especially display extreme irregularity, namely the blade base at the transition beti(Veen petiole and the lamina, and the transition between the lobed and unlobed blade sectors (indicated by arrows in fig 2e-f). The vein patterns of the parents are different in a number of characters. The vascular system of the blade of P. lanceolatum is more complex than that of P. th yssanolepis. In the former (except in the narrow tip) there are multiple layers of areoles; in the latter, there is normally but a single costal layer, with only rare anastomoses distally (cf figs 4a and b). The costal areoles of P. lanceolatum are narrow and parallel to the midrib; the medial and submarginal areoles tend to be somewhat more oblique. Those of P. thyssanolep is, however, are not only broader but they run at angles of 40-50° from the midrib. The marginal free veins in the entire-leaved species are projections from the outer circumferences of the marginal areoles and they are usually very short, less than 1 mm long. The marginal free veins in the pinnate species, on the other hand, are decidely longer, usually between 1-2.5 mm long. Also, the marginal vein tips in P. lanceolatum are only slightly expanded in contrast to those in P. thyssanolepis which WA GNER & WAGNER: A HYBRID POLYPODY FROM TH E NEW WORLD TROP ICS 127

''

,. ' �

FIGURE 1, Forms of Polypodium x leucosporum: a, "Type 1" (hypothetical formula, LL T, see tex t) , Santo Domingo, Fuertes 1964 (NY). b, "Type 3" (formula, LTT), Peru, Hu tch in son & Wrigh t660 (F). 128 FERN GAZETTE: VOLUME 11 PART 2 & 3 (1975)

FIGURE 2. a, Polyp odium lanceoletum (hypothetical formula LLLI:.., see tex t), Jamaica, field collection, Wagner 70101 (MICH). b, P. thyssenolepis (TTTT), field collection, 70 102. c, P. x leucosporum "Type 2" (LLTT), Jamaica, field collection, Wegner 70100. d, the same, greenhouse-grown simple forms like P. lenceolatum. e, the same, greenhouse-grown forms like P. thvssanolepis. f, the same, greenhouse giant forms. Arrows point to zones of blade showing max imum variation. WAGNER & WAG NER: A HYBRID POLYPODY FROM THE NEW WORLD TROPICS 129

THYSSANOLEPIS

25 � I I I • I • I I 20 ... I t I I I I Ill I

0 LANCEOLATUM

0 6 12 18 24

Blade Length in cm .

FIGURE 3. Lobation of leaf blades in relation to length of plants grown together under same greenhouse conditions. Base line: P. lanceolatum (no lobation at all). X-X: P. x leucosporum (numbers of lobes expressed as "full-sized" [dotted) and "abortive" [solid line) ). e-e: P. thyssan olepis. are conspicuously clavate with the terminations broadly rounded or ovate and associated with we ll developed, pocket-like, hydathodes (cf figs 5a and b). The fertile veins (i.e., the vein complexes subtending the soral receptacle) are also distinct in the parents. Under the sori of P. lanceolatu.m we find a complex reticulum, ovate in outline and composed of narrow bands of shortened tracheids. Subtending the sori of P. thyssanolepis, in contrast, there is usually a simple massive aggregation of shortened tracheids or at most a linear to lanceolate loop with a single cross-vein, as shown in figures 4 and 5. The sori of P. lanceolatum have a different location with respect to the over-all venation pattern. They are separated by 2-4 areoles from the midrib (except terminally ), but in P. th yssanolepis they are contained in the costal areoles adjacent to the midrib. In sterile and fertile vein characters, P. x leucosporum lies between its parents. From P. lanceolatum it receives the tendencies toward simplification of the blade, more numerous areoles, these more or less parallel to the midrib, and the more complex reticulate vein associated with the sari. From P. thyssanolepis it receives tendencies toward clavate hydathodal vein tips, as well as characteristics of the laminar scales to be described below. In all respects except one - the paraphyses - the hybrid appears to be a good intermediate. 130 FERN GAZETTE: VOLUME 11 PART 2 & 3 (1975)

FI GURE 4. Tracings of venation patterns of Jamaican polypod ies: a, Po/ypodiu m Janceolatum; b, P. th yssanolepis; c, P. x leucosporum. Right side: Fertile veins much enlarged to show simpler forms of each taxon.

The scales of the abaxial blade surface nicely distinguish the parents and the cross. In P. /an ceo/atum (fig 6a ) the laminar scales are tiny, mostly 0.3-0.4 mm in diameter and round in outline. The "fringe" of th i nner-walled, pale cells is relatively narrow, only about equal in width to the diameter of the dark central area of th ick-walled cells. The scales are scattered remotely over the blade surfaces at distances averaging one or more millimeters. In P. thyssanolepis (fig 6b) the scaies are much larger, being 2-5 times greater in over-all dimensions and mostly lanceolate in outline with an elongate tip. The "fringe" of pale cells is larger and the cells are elongated. Also the scale margin is conspicuously fimbriate. The scales are closer, covering between one-third and one-half of the blade surface and many of them overlap. The scales of P. x leucosporum (fig 6c-d ) are transitional between these extremes in size, form, and distribution. This proves useful in identification : occasional simple fronds of P. x leucosporum (e.g., fig 2d) that might be confused with P. lanceolatum can be distinguished by their larger, more lanceolate and more densely distributed scales. The sori of P. /anceolatum are distinguished by their large size, ov'ate outl ine, and presence of conspicuous peltate paraphyses. The paraphyses are trichomes like those of the lamina but they are borne on long stalks and appear above the young sporangia as a more or less continuous "roof'. As the sorus becomes older, the paraphyses separate and fall off. The flattened terminal "umbrellas" have bl�ckened centers with clathrate secondary wa lls and broad tan to white margins made up of cells with unmodified wa lls. Such paraphyses are absent in P. thyssanolepis. Remarkably they are also apparently absent in P. x leucosporum. At first we thought that P. thvssano/epis and P. x /eucosporum were without WAGNER & WAGNER: A HYBRID POLYP ODY FROM THE NEW WORLD TROPICS 131

FIGURE 5. Photographs of cleared fertile veins: a, P. lanceolatum, b, P. thyssanolepis; c, P. x /eucosporum, simple form on left, more complex on right. paraphyses of any sort, but we !ater discovered that there are receptacular paraphyses among the sporangia! stalks hidden below the sporangia! capsules. They are very sparse and difficult to find in soral preparations. They were later discovered in P. lanceolatum as we ll. These paraphyses are narrow filaments of 2-9 cells in length. In P. lanceolatum and the hybrid they are normally between 150 and 300pm long with 132 FERN GAZETTE: VOLUME 11 PART 2 & 3 (1975)

• c

FIGURE 6. Laminar scales; a, Po/ypodium lanceo/atum; b, P. th yssanolepis; c, P. x leucosporum.

3-5 cells; in P. thyssano/epis they are shorter, 100-150 ,um, with only 2 cells.

CYTOLOGY Studies of the chromosomes proved valuable in interpreting our material. In his discussion of P. lanceo/atum in Jamaica, Walker (1973) noted tetraploid and pentaploid forms. The pentaploid had irregular meiosis, and two of the three tetraploids also had irregular me iosis. Presence of both 4x and 5x forms indicates the existence of 2x forms. Chromosome counts of our specimens from Jamaica gave 74 pairs and a somatic count of ea. 148. Our P. th yssano/epis also gave 74 pairs (fig 7c). The hybrid polypody showed irregularity in meiosis, with uneven nucleus formation and excluded chromosomes. All figures contained, however, a high WAGNER & WAGNER: A HYBRID POLYPODY FROM THE NEW WORLD TROPICS 133

�, • ..,. ... � '• , r • • • J , . • ... ' • -� . ., " ,\ .J -.. f t I - • • . . .. • *' I " f . � • t,. � ,..� * .... '• ,. • ) "' \ ) ,;-- I: t • C! J , .. .JI( ' ( • 'l ;,t I • • " ) ' • lr �.- , . "'< , " • { .. • , .. )Y. • , ) ... � \ • \ ! • , , � ... ' " ..._ ·� ,)'

a b

.. • ..

FIGURE 7. Chromosomes squashes at meiotic metaphase. a, Po/ypodium x /eucosporum, 59 11 + 301. b, the same, 63 11 + 22 1. c, P. th yssanolepis. incidence of pairs at meiotic metaphase, suggesting either compatibility between homoeologous chromosomes or autosyndesis or both. Five sporocytes examined in detail revealed the following patterns: 45 11 +58 I 51 11 +461 59 11 +30 I 63 11 + 22 I 65 11 + 18 I Aver�ge 56.6 11 + 34.8 I

From this evidence we conclude that at our locality in Jamaica, P. x leucosporum arose as a sterile cross of tetraploid forms of the parental species. By virtue of its 134 FERN GAZETTE: VOLUME 11 PART 2 & 3 (1975) vigorous creeping rhizome growth, the hybrid has formed a large colony. The spores are abortive, and presumably unable to reproduce the plant.

DISCUSSION Hybrid polypodies more or less like the one described here have been taken sporadically from the Greater Antilles and Mexico to Peru. They seem to have been collected particularly often in Colombia and Venezuela. There are possibly genetically 'different forms of P. x leucosporum and indeed it is likely that some of the polypodies resembling it may have resulted from other species 'combinations (e.g., P. bartlettii )Neatherby, 1935). There seem to be three forms of P. x leucosproum, as was first suggested by Vareschi (1969: 193). We hereby designate these as follows : 'Type 1'' (fig 1a), closer to P. lanceolatum, the blade linear-lanceolate, remotely scaly, with shallow lobes or undulations; "Typ e 2" (fig 2c) intermediate, the blade broader, more scaly, with equal numbers of shallow lobes and long lobes, and a long tip (like the plants described herein); and "Type 3" fig 16), closer to P. thyssano/epis, the blade oblong-lanceolate, densely scaly, with many fully developed lobes, few shallow lobes, and a relatively shorter tip.* As both parents have diploid tetraploid forms (Walker 1973, Evans 1963, present paper), they can make different contributions to the hybrid. Perhaps, then, the followi ng scheme may explain the variations where the superscripts are added to indicate pairing factors. P. lanceolatum (L) P. thyssanolepis (T) P. x leucosporum L1 L1L2L2 (4x) TT (2x)---L 1 L2T (3x)" Type 1" LL (2x) X TT (2x)--- LT (2x) "Type 2" Dip. form L 1 L 1 L2L2 14x ) X T1T 1T2T2 (4x ) --- L 1 L 1 T2T2 (4x) "Type 2" Tet. form LL (2x) X T 1T 1T2T2 (4x) --- LT 1 T2 (3x) "Type 3"

Of special interest is the fact that P. x leucosporum is the cross of a taxon commonly placed in subgenus or genus P/eopeltis and a taxon commonly placed simply in Po/ypodium or in subgenus or genus Marginaria. Po/ypodium lanceolatum clearly conforms to P/eope/tis, having peltate paraphyses and "veins anastomosing freely and irregularly with included veinlets; sori borne at the union of several veinlets" (from Copeland, 1947, p. 183). On the other hand, P. th yssanolepis fits Copeland's description (p. 180) of one of the common conditions in Po/ypodium: "The areolae usually form one row on each side of the costa, with always a single free vein in each areola. The sori are terminal on these free included veinlets" (I.e.). Without question, if these species can hybridize and combine so-called "generic" characters then there is room to doubt if tt]ey are taxonomically tenable.

•Examples· of "Type 2" are most common in herbaria. Of the others, the following seem to be good examples: "Type 1" - HAITI: Base of Guimbi Galeta, Holdridge 1770 (US); Massif de Cahos, Ekman H. 3478 (US). SANTO DOMINGO: Cordillera de Neyba, Ekman H- 13489 (US); Calo de Maco, F.uertes 1964 (NY). VENEZUELA:State of Merida, Mucuruba, Gehriger 304 (US, NY - this paper, fig. 1, left). "Type 3" - COLOMBIA: Dep. Antiogu ia, Medelliu, Daniel 33 10 (US); Auf Mauera in Tasto, Lehmann 580 (US). PERU: Dep. Huanaco, P.rov. Pachitea, La Molina, Asplund 73662 (US); Dep. Piura, Prov. Huancabamba, Hucchinson and Wrigh t 660 1 (F - Fig. 1, right, NY, US). We have also noticed several specimens labelled as "P. leucosporum " Which we think may represent combinations of other species as patents, including the following: VENEZUELA: Monogas, Cerro Negro, Steyermark 62109 (F, NY, US). COLOMBIA: Zipaqu ire, Huertas and Camargo in 1942 (US). WAGNER & WAGNER: A HYBRID POLYPODY FROM THE NEW WORLD TROP ICS 135

ACKNOWLEDGEMENTS Th is research was supported in part under National Science Foundation Grant . . GB-309 18x, "Th e Evolutionary Characters of Ferns". We. are grateful for the aid of William E. Hammitt of Albion College, Michigc.n, and George R. Proctor of the Institute of Jamaica. We also thank the curators of the Royal Botanical Gardens at Kew, the Field Museum (Chicago). the U.S. National Herbarium, and the New York Botanical Garden, for graciously lending us herbarium collections.

REFERENCES COPE LAND, E.B. 1947. Genera filicum. Chronica Bota nica, Wa ltham, Mass. EVANS, A.M. 1963. New chromosome observations in the Polypodiaceae and Grammitidaceae. Ca ryologia 16: 671-677. LANG, F.A. 1971. The Polypodium vulgare complex in the Pacific Northwest. Madron7f21: 235-254. LE LLINGER, D.B. 1968. Notes on Ryukyu ferns. Amer. Fern J. 58: 155-158. VARESCH I, V. 1969. Flora de Venezuela. Helech os. Vol. 1: 473-1033. WAGNER, W.H., Jr. 1962. Irregu lar morphological development in hybrid ferris. Ph ytomorphology 12: 87-1 00. WALKER, T.G. 1973. Additional cytotaxonomic notes on the pteridophytes of Jamaica. Trans. R. Soc. Edinb. 69: 109-135. WEATH ERBV, C.A. 1935. On certain Mexican and Central American Ferns. Amer. Fern J, 25: 52-59.

FERN GAZ. 11(2 & 3) 1975 137

ASPID ISTES THOMAS! I-A JURASSIC MEMBER OF THE THELYPTERIDACEAE

J.D. LOVIS Department of Plant Sciences, University of Leeds, LS2 9JT.

Harris (1973) has drawn our attention to the neglect of late Mesozoic and Tertiary fern fossils, pointing out that these have usually received on ly rather superficial study and "glib determinations deserve and get little attention" (I.e., 43). The fault is however clearly mutual, insubstantial work by palaeobotanists being matched by disinterest and diffidence on the part of students of living ferns. To be able to suggest, as does Blasdell ( 1963: 30) in his monograph on Cystopteris that the disjunct distributional pattern of this genus can be explained if it is postulated that it evolved during the Carboniferous period is only possible if no note is taken of fossil evidence. A common and more understandable mistake is too ready an acceptance of palaeobotanical records at their face value, or even sometimes at rather more than their face value. For example, it is unfortunate that an otherwise excel lent and valuable analysis of the Polypodiales by de la Sota (1973) is marred by the uncritical assumption that a Triassic record given the name Polypodites indicates the existence of Polypodiaceae (sensu stricto ) at that time. In consequence, he suggests (I.e., 242) that "The Polypodiaceae would be at least as old as the Dipteridaceae", a conclusion which is certainly contrary to the actual fossil evidence. The lack of liaison between palaeobotanists and pteridologists can th us lead the latter into regrettable error. However, Or R.E. Holttum constitutes an extremely distinguished exception to the generally rather indifferent attitude of students of modern ferns towards palaeobotany. Some readers of the Fern Gazette may not know that more than fifty years ago, at the outset of his career, Holttum worked as a palaeobotanist with Seward, and published with him studies on Mesozoic and Tertiary floras, from Rhodesia, Ceylon and Mull (Seward & Holttum, 1921 ; 1922; 1924 ). Furthermore, he had the surely unforgettable experience of assisting Seward in the collection of Cretaceous plant fossils in Greenland (Holttum, 1922; Seward, 1926). Stimulated by Prof. Harris's final comments in his address to the 1972 Symposium at the Linnean Society, and ignoring the warning of a well-known proverb about fools rushing in where angels fear to tread, I have recently ( Lovis, in press) attempted to survey the fossil records of early "polypodiaceous" ferns. In consequence, I endorse Harris's opinion that very, very. few Mesozoic fossifs can be confidently accepted as "polypodiaceous" ferns. Only one such fossil is known from Jurassic rocks. This is Aspidistes · th omasii Harris*. Holttum has discussed th is fossil, albeit briefly, on three occasions (Holttum 1963 : 65-67; 1971 : 18; 1973b: 177 ). However, since his opinion of this fern appears to be as yet unappreciated by palaeobotanists (I have yet to see it mentioned in print outside his own publications), I take advantage of this very appropriate opportunity to draw attention here to his comments on this very significant fossil. Aspidistes th omasii was very well described by Harris (1961 ) on the basis of material collected by Hamshaw Thomas from the famous Gristhorpe Bed in Cayton Bay, Yorkshire Harris placed it, albeit with reservations, in the family Aspideae (sic. ) *Another species of Aspidistes (A . beckerii Larch) has been described from Jurassic rocks in Israel (Larch 1967) but th is fossil is quite ev idently entirely distinct from A. th omasii, and in my opinion is only very doubtfu lly of "polypodiaceous" character at all. ( Lovis, in press) . 138 FERN GAZETTE: VOLUME 11 PART2 & 3 (1975)

However, Holttum states (1963: 67 ) that "A sp idistes looks like an early The/yp teris ': Of course, Holttum is uniquely well qualified to pronounce on this particular subject, not merely by reason of his early experience as a palaeobotanist but because of his unrivalled knowledge of the Thelypteridaceae, derived from detailed and extensive studies on the family (Holttum, 1954; 1969; 1971; 1973b ). in consequence of which he stands without equal as an authority on this group. To those whose experience of Thelypteridaceae is limited to British taxa, the justification of associating with the Thelypteridaceae Aspidistes th omasii, possessing as it does "repeatedly pinnate" fronds with conspicuous and persistent round indusia, may, appear rather less than obvious, since our three native species, Phegopteris connectilis (Michx.) Watt, Oreop teris limb osperma (All.) Holub and Th elyp teris palustris Schott, all have fronds which are only bipinnatifid, and are either exindusiate, or bear very delicate evanescent indusia. The shape of the pinnules of A th omasii resembles those of species of the tropical and sub-tropical genera Pseudophegop teris and Macrothelypteris. The total size of the frond of A th omasii is not known, but it may have been large, since the fossil fragments can be well-matched by pinnae of the Malaysian species M. polypodioides (Hook.) Holttum (= The/ypteris leucolepis (Presl ) Ching). They are also close in form to th e better known and more widespread M. to rresiana (Gaud.) Ching (= T. uliginosa Ching). In Macro thelyp teris, the indusia are persistent but small, but large and persistent indusia, comparable to those of A thomasii, are present in some species of Coryphopteris, a genus which includes forms which "may represent the most primitive extant members of the family" (Holttum, 1973b: 177). Furthermore, the lamina and indusia of A. th omasii bear unicellular glands which are very simi far to the glands of C. viscosa (Bak.) Holttum. A notable feature of A th omasii is the possession of trilete spores. Holttum has pointed out (1971 : 29) that "If, as I believe, Th elypteridaceae have a relationsHip to Cyathea, trilete spores must have been primitive in the family". The spores of modern Thelypteridaceae are almost without exception monolete, but one small living genus, Trigonospora, does indeed possess trilete spores, even though it "in other respects is not evidently primitive". (Holttum, 1971: 29). Thus, though no living thelypteroid possesses all the characteristics of Aspidistes th omasii, there is no character of this fossil which cannot be matched somewhere amongst modern Thelypteridaceae, and there appears to be no good reason why it should not itself be regarded as a member of this family. lt follows that individual primitive thelypteroid characteristics (i.e., those seen in A thomasii) have been conserved in certain different living genera, namely Coryphopteris, Macrothelyp teris (and Pseudophegopteris ) and Tr igonospora. Holttum has long since (1947, 1949) argued that the thelypteroids are quite distinct from the dryopteroid ferns, an opi nion subsequently clearly endorsed by the cytological evidence (Manton Bt Sledge, 1954; Walker, 1973 ). More recently, with the re-unification of the tree-ferns within a single family (Holttum & Sen, 1961 ). it became evident that the thelypteroids and dryopteroids could not be as widely separate in their points of origin as Holttum had earlier supposed, but he has emphasised (Holttum 1969: 5, 6; 1973a : 5) that he still regards these two groups as having arisen from quite distinct sources within the Cyatheaceae, the Thelypteridaceae being related to Cyathea, whereas the origin of the dryopteroids- is traced to the vicinity of the Thyrsopteridoideaet. tFollowing Holttum & Sen (1961), Culcita (with its possible fossil relatives, e.g. Coniop teris spp., cf. Harris, 1961; Holttum 1965: 63 ) is here included under Thyrsopteridoideae with in Cyatheaceae, and therefore excluded from Dennstaedtiaceae, sensu. lato, contrary to Holttum 1947, 1949 ). LOV IS: ASPIDISTES THOMASII 139

lt follows that the detection of a thelypteroid fern in the Jurassic can tell us nothing regarding the emergence of the main radiations of ferns with superficial and indusiate sari. There is in my opinion as yet no convincing fossil example of Dennstaedtiaceae (sensu lata, and therefore including the dryopteroids, e.g., Holttum 1947, 1949) known from the Mesozoic period. The earliest fossil known to me acceptable as a dryopteroid fern is Dryopteris meeteeseana Brown, of Palaeocene age (Brown, 1962), though even this example is perhaps more likely to be a tectarioid ( Lovis, in press). The evidence is as yet far too fragmentary for conclusions to be drawn, but it is possible that the distinctness of the Thelypteridaceae, already attested by cytology, may in time be further confirmed by this family proving to be of significantly greater age than the dryopteroid ferns ..

REFERENCES BLASDE LL, R.F. 1963. A monographic study of the fern genus Cystopteris. Mem. To rrey bot. Clu b, 2 1: 102. BROWN , R.W. 1962. Paleocene flora of the Rocky Mou ntains and Great Plains. U.S. Geol. Survey Pro f. paper, 375: 1-119. HAR RIS, 'T.M. 1961. The Yorkshire Jurassic Flora, I: Th allophy ta-Pteridophyta. London, British Museum (Nat. H ist.). HARRIS, T.M. 1973. What use are fossil ferns? In A.C. Jermy et al. (eds. ), The Phylogeny and Classification of the Ferns. Bot. J. Linn. Soc. 67, Suppl. 1: 41-44. HOL TTUM, R.E. 1922. The vegetation of West Greenland. J. Ecol. 10: 87-108. HOLTTUM, R.E. 1947. A revised classification of the leptosporangiate ferns. J. Linn. Soc. (Bo t.) 53: 123- 158. HOLTTUM, R.E. 1949. The classification of ferns. Bioi. Rev. 24: 267-296. HOLTTUM, R.E. 1954. Flora of Malaya, 11 : Ferns of Malaya. Singapore, Government Printing Office. HOL TTUM, R. E. 1963 Cyatheaceae. Flora Malesiana, Ser. 11. Pteridophym · 1 (2 ): .65-1 76. Grongingen. HOLTTU M, R.E. 1969. Studies in the fami ly lhelypteridaceae. The genera Ph egop teris, Pseudophegopteris, and Macro thelypteris. Blumea 17: 5-32. HOLTTUM, R.E. 1971. Studies in the family Thelypteridaceae Ill. A new system ofgenera in the Old World. Blumea 19: 2-52. HOLTTUM, R.E. 1973a Posing the Problems. In A.C. Jermy et al. (eds.), The Phylogeny and Classification of the Ferns. Bot. J. Linn. Soc. 67, Suppl. 1: 1-10. HOLTT UM, R.E. 1973b. The family Thelypteridaceae in the Old World. In A.C. Jermy et al. (eds.), The Phylogeny and Classification of the Ferns. Bot. J. Linn. Soc. 61, Suppl. 1: 173-189. HOLTTUM, R.E. & SEN, U. (1961). Morphology and classification of the tree-ferns. Ph ytomorphology 11: 406-420. LORCH, J. 1967. A Jurassic flora of Maktesh Ramon, Israel. lsreal J. Bot 16: 131-155, pi. 162-180. LOV IS, J.D. 197?. Evolutionary patterns and processes in ferris. Adv. bot. Res. 4�(in press). MANTON, I. & SLEDGE, W.A. 1954. Observations on the cytology and taxonomy of the pteridophyte flora of Ceylon. Phi/. Trans. Roy. Soc., Ser. B, 238: 1 27-185. SE WA RD, A.C. 1926. The Cretaceou s plant·bearing rocks of Western Greenland. Phi/. Trans. Roy. Soc., Ser. B, 215: 57-1 76. SEWARD, A.C. & HOLTTUM, R.E. 1921. On a collection of fossil plants from South Rhodesia. S. Rhodes, Ceol. Surv. Bull. 8: 39-45. SEWARD, A.C. & HOLTTUM, R.E. 1922. Jurassic plants from Ceylon. Quart. Journ. Geol. Soc. 78: 271 -277. SEWARD, A.C. & HOLTTUM, R.E. 1924. Palaeobotany (=Ch. IV. Tertiary plants from Mull). In Bailey, E.B. et al., Tertiary and Post-Tertiary Geology of Mull, Loch Aline and Oban, pp. 67-90. Mem Geol. Surv. Sco tland, Edinburgh, HMSO. SOTA, E.R. de la. 1973. On the dassification and phylogeny of the Polypodiaceae. In A.C. Jermy et al. (eds.) , The Phylogeny and Classification of the Ferns. Bot. J. Linn. Soc. , 67, Suppl. 1: 229-244. 140 FERN GAZETTE: VOLUM E 11 PART 2 & 3 (1975)

WA LKER, T.G. 1973. Evidence from cytology in the classification of ferns. In A.C. Jermy et al. (eds.), The Phylogeny and Classification of the Ferns. Bot. J. Linn. Soc., 67, Suppl. 1: 91-1 10. FE RN GAZ. 11(2 & 3) 1975 141

A NEW GENERIC SEQUENCE FOR THE PTERI DOPHYTE HERBARIUM

J.A. CRABBE and A.C. JERMY Botany Department, British Museum (Natural History), London SW7 5BD and J.T. MICKEL The New York Botanical Garden, Bronx, New York 10458

ABSTRACT In a preamble to Index Fil icum, Christensen ( 1906) listed a classification scheme of 136 genera which was based on that of Engler & Prantl (1898-1902) and which he considered worthy of recognition. This arrangement, already accepted in principle by the older herbaria, was accepted and adopted for the nex t 30 to 40 years wherever curators required a systematic (phylogenetic) order for the ferns. In 1938 Christensen revised his 1906 classification and there followed at least ten other proposals by other workers putting toward phylogenetic schemes. The scheme presented here is the resu lt of discussions with a large number of pteridologists in the course of rearranging the pteridophytes at the British Museum and New York Botanical Garden.

INTRODUCTION The diversity of opinion and lack of consensus on a taxonomic treatment of the ferns causes problems for all devising a useable herbarium arrangement or for that matter for those wishing to arrange fern data systematically in any other data bank, e.g., a computer, visual card index, etc. The authors have been wrestl ing with the problem of the most suitable arrangement for the large fern col lections at thei r institutions and had to answe r the fo llowing questions: should the herbarium curator follow a traditional treatment until the dust of disagreement settles; or is consensus just around the corner and can we devise a new arrangement. The history of fern classification from that of Smith (1793, 1810) onwards has been reviewed by Pichi Sermolli (1 973). Whilst some early curators used Presl (1836, 1845), possibly the most widely accepted, if not the most important classification, was that used by Engler & Prantl (1898-1 902 ). The Polypodiaceae s.l. account (pp 139-339) was written by Diels who devised a classification of nine tribes based on sorus structure. The Engler and Prantl system was adopted by Christensen in the preamble to his Index Filicum (Christensen 1906). For the next thirty years writers and cu rators alike followed this arrangement in spite of its unnatural sequence of families and misconception and incorrect systematic position ·of some genera. Diels attempted to show fern relationships as he saw them, but in light of our present knowledge, his is now quite misleading. For example, the genera Dennsta edtia, Hypo/epis and Pteridium are situated far apart in the Diels scheme yet today we know that they are closely related; Dryopteris includes most of the thelypterids along with Ctenitis and other genera generally recognised as distinct today; the woodsioid ferns are apart from other athyrioids and are placed near the tree ferns; Nephrolepis is separated from the other davallioids. In 1938 Christensen revised his 1906 outline (slightly changed by him in 1934) in view particularly of the work of Bower (1923-28). Since 1940 no fewe r than ten botanists have proposed phylogenetic schemes for the ferns (Ching 1940, Dickason 1946, Copeland 1947, Holttum 1947, 1949, 1973a, Pichi Sermolli 1958, Mehra 1961, Sporne 1962, Wagner 1969, Nayar 1970, Mickel 1973, 1974). As in any group involving the activity of many workers, 142 FERN GAZETTE: VOLUME 11 PART 2 & 3 (1975)

there resu lts a great diversity of opinion on the taxonom.ic arrangement within that group. This wide assortment of treatments includes examples of strong taxonomic amalgamation to extreme fragmentation. The most severe disagreement at the family level centres around the Polypodiaceae which has been maintained traditionally as a single family or divided into as many as 33 families (Ching 1940). A significant contribution to a phylogenetic arrangement of the Polypodiaceae· was made by Holttum (1947, 1949 ). At the same time Copeland (1947) produced his Genera Filicum, again presenting a new arrangement in places misleading with unnatural groupings. Furthermore, the past 25 years have been marked by intensive research on the ferns and fern allies. Much of this was triggered by Manton's classic cytological work (1950),but studies have proceeded in many other directions, including anatomy, chemistry, genetics, physiology, development and palynology. Some of these studies _ have merely used ferns as research tools (Mickel 1971 ) but others have contributed substantially to a better understanding of the relationships of ferns. These advances have been we ll summarised in three recent symposia (Tr\ion 1971, Jermy, Crabbe & Thomas 1973, Tay lor & Mickel 1974). We should like to suggest that an arrangement of taxa in an herbarium might take any form, and selection of a system should be dictated by its primary users. Thus, a small regional herbarium where all the genera are known might find it most practical to maintain the ferns alphabetically by genus or according to a local Flora or monograph; other herbaria may wish to keep a well-established arrangement simple because the primary users of the herbarium know where everything is placed. lt is our feeling that a large collection that is used daily for identification, floristic, or monographic purposes should be arranged to reflect relationships in the best way possible. Proximity of related plants is essential to facilitate consultation during identification.· lt enables a floristician or monographer to better see the boundaries between genera and to find valuable materials filed in related groups. An alternative to a phylogenetic scheme is an alphabetical system. lt has the advantage of ease of filing by untrained personnel and the ease of finding genera by visitors to the herbarium. However, the system is fixed in a most awkward manner. Generic names cannot be readily brought up·to-date without major reorganisation to the herbarium or an elaborate cross-reference system. Related genera cannot readily be compared since the alphabetical system separates without bias all genera from their relatives. Such a system may be useful in certain herbaria where there is no pteridologist or, as suggested above, where the col lection is very small. In preparing the followi ng system we have tried to incorporate the most recent evidence bearing on phylogeny, including gross morphology, cytology, anatomy, gametophytes, palynology, and chemistry. We have conferred with specialists of as many groups as possible to get their opinions on particular problems and with leading pteridologists throughout the world. We wish to thank all who have provided us with valuable information and suggestions but emphasise that we alone are responsible for any �rrors or misconceptions that time may show to have crept in. We feel that the family concept is still in a state of flux and requires considerable monographic work at the species and genus level before further conclusions should be drawn. Family and subfamily headings are given solely to indicate the general systematic layout. As Christensen (1938: 52 5) pointed out· "the linear sequence of families and genera moreover gives no clear idea of the highly interwoven phylogeny of recent ferns and it is perhaps impossible to demonstrate to others the reticulated phylum as it appears to the trained taxonomist", lt is appreciated that the sequence of these major units can vary according to the phylogenetic route taken. We have done CRABBE, JERMY & MICKEL: A NEW GENERIC SEQUENCE 143

ou r best to place related groups of genera in justaposition.

Numbering and layout of the list Mostly we have followed the concepts of higher taxa generally agreed by pteridologists but feel that a hierachical layout as proposed here shows relationships more clearly.

For this reason we have adopted the oldest family name available, as demanded by· the Code, and have formal ised certain subfamily names where necessary. Each family and subfamily unit is separated by a solid line. Where we think such units could be divided, a broken line .has been inserted. As the emphasis in this treatment is on the genus, the famil ies and subfamilies are not numbered. The numbers are thus consecutive throughout the list; new genera or curatorial units may be added as a number following a point, i.e. using an extended decimal system. Certain of the generic names are indented. These are (i) genera which have been segregated, often on a regional basis (e.g. Ching 1954) and with an incomplete knowledge of the group, and which we feel are in need of revision; (ii) gen�ra now regarded as giving a mo re constructive picture of the classification· of a major group hitherto (and by some pteridologists, still) regarded under a single genus (e.g. Th e/yp teris, Cyathea ); (iii) genera morphologically distinct and traditionally accepted as separate but which recent studies have shown to be genetically close (e.g. Asplenium); or (iv) where we feel further studies wi ll reveal closer associations or new generic limits with in the group. The indented names have been given a full number in the sequence of genera to avoid compl icating the system. Should certain indented (or non-indented for that matter) genera be not accepted in a system the numbers can be omitted or placed in brackets (e.g. 288 (289-294) Asplenium ) or, if the species group (genus) is accepted at a curatorial level (as subgenus or section) within the aggregate genus, the number may follow in brackets e.g. 288 (289) Asplenium (sect. Camp tosorus). lt should be noted that these names are legitimate generic names and not subgeneric and if the latter rank is used there may be a prior name at that rank; e.g. in the case of Ctenop teris if placed as a subgenus of Grammitis the name Cryptosorus must be used. We have listed the genera_in the followi ng assemblages : 1. Fern allies. 2. Eusporangiate ferns (Ophi oglossales, Marattiales, Osmundales) and Plagiogyriales. 3. Sch izaeoid ferns � adiantoid ferns Note: Initially we tried to place all the traditional Polypodiaceae sens. fat. together for those curators who do not wish to make sweeping changes. However we believe - and there is an increasing opinion accepting - the adiantoid ferns to be a distinct group arising from schizaeaceous stock. 4. The filmy ferns and related groups Note: generic concepts in the Hymenophyllaceae wi ll remain in a state of flux until a large-scale mo nograph is carried out. We have followed Morton (1968) in arrangement and taxonomy excepting that we have accepted his Sections at generic rank for those wi shing to break down the large umbrella genera of Trichomanes and Hymenophyllum. In those five cases where no generic name exists . we have .used the Sectional name followed by (p ro sectione) rather than make a new combination and status for a Section which may on further study remain at that, or even be reduced in, rank. 5. Gleichenioid ferns ------7 polypodioids and grammitid ferns, by way of Matonia, Cheiropleuria and Dipteris. 144 FERN GAZETTE: VOLUME 11 PART 2 & 3 (1975)

6. Proto-cyatheoid and cyatheoid ferns � dennstaedtioid � thely­ pterid � asplenioid ferns, the _" Aspidiales" and t�e _blechnoid ferns. 7. Hydropterides Note: We place the water ferns here not because we feel they are related to each other, but out of convenience. They are quite distinct and there is no firm data to justify placing them elsewhere.

SYSTEMATIC LIST OF PTE RIDOPHVTE GENERA

Psilo1aceae 001 Psilotum 002 Tmesipteris

Lycopodiaceae 003 Lycopodium 004 Lycopodiella 005 Diphasium 006 Huperzia 007 Phylloglossum

Selaginellaceae 008 Selaginella lsoetaceae 009 lsoetes 010 Stylites

Equisetaceae 01 1 Equ isetum

Oph ioglossaceae 012 Botrychium 013 Botrypus 014 Sceptridlum 015 Helminthostachys 016 Ophioglossum 01 7 Cheiroglossum 018 Rhizoglossum

Marattiaceae 019 Ang iopteris 020 Archangiopteris 02 1 Protomarattia 022 Marattia 023 Macroglossum 024 Danaea 025 Christensenia

Osmundaceae 026 Osmunda 027 Osmundastrum 028 Plenasium 029 Todea 030 Leptopteris

Plagiogyriaceae 031 Plagiogyria

Schizaeaceae 032 Schizaea 033 Actinostachys 034 Lygodium 035 Anemia 036 Mohria

Parkeri aceae 03? Ceratopteris

Platyzomataceae 038 Platyzoma

Adiantaceae: Adiantoideae 039 Actiniopteris CRABBE, JERMY & MICKEL: A NEW GENERIC SEQUENCE 145

Adiantaceae: Adiantoideae (c ontinued) 040 Afropteris 041 Ochropteris 042 Anopteris

043 Onychium 044 Cryptogramma 045 Llavea

046 Neurosoria

047 Cheilanthes 048 Adiantopsis 049 Aleuritopteris 050 Aspidotis 051 Cheiloplecton 052 Mildella 053 Notholaena 054 Argyrochosma (pro sectione) 055 Sinopteris 056 Negripteris

057 Pellaea 058 Doryopteris 059 Ormopteris 060 Safford ia

061 T rachypteris

062 Anogramma 063 Pityr�ramma 064 Trismeria 065 Pleurosoriopsis

066 Hemionitis 067 Gymnopteris 068 Bommeria 069 Paraceterach

070 Pterozonium 071 Syngrammatopteris 072 Jamesonia 073 E rlosoru s 074 Nephopteris

075 Syngramma 076 Craspedodictyum 077 Toxopteris 078 Austrogramme

079 Taenitis 080 Platytaenia 081 Holttumlella

082 Coniogramme

083 Aspleniopsis 084 Rheopteris 085 Cerosora

086 Adiantum 146 FERN GAZETTE: VOLUME 11 PART 2 & 3 (1975)

Adiantaceae: Vittarioideae 087 Antrophyum · 088 Polytaenium 089 Scoliosorus 090 Anetium 091 Hecistopteris 092 Ananthacorus 093 Vittaria 094 Monogramma 095 Vaginularia

Pteridoideae 096 Pteris 097 ldiopteris 098 Copelandiopteris 099 Neurocallis 100 Acrostichum

Loxsomaceae 101 Loxsoma 102 Loxsomopsis

Hymenophyllaceae 103 Hymenoglossum 104 Serpyllopsis 106 Rosenstockia 106 Hymenophyllum 107 Buesia 108 Meringium 109 Eupectinum (pro sectional 110 Myriodon 111 Sphaerocionium 112 Apteropteris 113 Craspedophyllum 114 H emicyatheon 115 Mecodium 116 Cardiomanes 117 Trichomanes 118 Lacosteopsis (pro sectional 119 Crepidomanes 120 Polyphlebium 121 Reediella 122 Abrodictyum 123 Pleuromanes 124 Gonocormus 125 Selenodesm ium 126 Davalliopsis 127 Cephalomanes 128 Callistopteris 129 Nesopteris 130 D idymoglossum 131 M icrogonium 132 Lecanolepis 133 Achomanes (pro sectional 134 Neuromanes 135 Odontomanes 136 Lacostea 137 Trlgonophyllum (pro sectional 138 Homoeotes 139 Feea 140 R agatelus CRABBE, JERMY, MICKEL: A NEW GENERIC SEQUENCE 147

Hymenosphyllaceae (continued) 141 Acarpacrium (pro sectione)

Hymenophyllopsidaceae 142 Hymenophyllopsis

Stromatopteridaceae 143 Stromatopteris

Gleicheniaceae 144 Gleichenia 145 D iplopterygium 146 Sticherus 147 Dicranopteris 148 Acropterygium

Matoniaceae 149 M atonia 150 Phanerosorus

Che irop leu riaceae 15 1 Cheirop leurla

D ipteridaceae 152 Dipteris

Polypodiaceae: Drynarioideae 153 Drynaria 154 Photinopteris 155 Merinthosorus 156 Aglaomorpha 157 D ryostachyu m 158 Holostachyum 159 Drynariopsis 160 Pseudodrynaria 161 Thayeria

Platycerioideae 162 Platyc�rium . -- - -- 163 Pyrrosia 164 Saxiglossum 165 Drymoglossum

M icrosorioideae 166 Microsorum 167 Dendroconche 168 Phymatosorus 169 Lecanopteris 170 Colysis 171 Podosorus 172 Diblemma 173 Leptoch ilus 174 Para leptoch i Ius 175 Dendroglossa m5 Christiopteris 177 Pycnoloma 178 Grammatopteridium 179 0 leandropsis 180 Holcosorus 181 Crypsinus 182 Selliguea 183 Arthromeris 184 Polypod iopteris

Pleopeltoideae 185 Pleopeltis 186 M icrogramma 187 . Anapeltis 188 Craspedaria 189 Solanopteris 190 M arginariopsis 148 FERN GAZE 1TE:VO LUME 11 PART 2 & 3 (1975)

Polypodiaceae: Pleopeltoideae (continued) .1 91 Neurodium 192 Neolepisorus 193 Lemmaphyllum 194 Drymotaenium 195 Paragramma 196 Betvisia 197 D icranoglossum 198 Neocheiropteris

Polypodia ideae 199 Niphidium 200 Pessopteris 201 Campyloneu rum 202 Anapausia 203 Dictymia 204 Phlebodium 205 Synammia 206 Polypodium 207 Goniophlebium 208 Thylacopteris 209 Marginaria

Grammitidaceae 210 Grammitis 21 1 Xiphopteris 212 Ctenopteris 213 Calymmodon 214 Acrosorus 215 Amphoradenium 216 Prosaptia 21 7 Glyphotaenium 218 Oreogrammitis 219 Nematopteris 220 Scleroglossum 221 Cochlidium ------222 Hyalotricha 223 Loxogramme 224 A narthropteris

Metaxyaceae 225 Metaxya

Lophosoriaceae 226 Lophosoria

Cyatheaceae 227 Cnem ldaria 228 Cyathea 229 Trichopteris 230 Nephelea 23 1 Alsophila 232 Sphaeropteris 233 Dicksonia 234 Cystodium

Thyrsopteridaceae 235 Thyrsopteris ------236 Culcita ------237 Cibotium

Dennstaedtiaceae: Dennstaedtioideae 238 Dennstaedtia 239 Microlepia 240 Oenotrichia 241 Leptolepia - -- CRABBE, JERMY & MICKEL: A NEW GENERIC SEQUENCE 149

Dennstaedtiaceae: Dennstaedtioideae (c ontinued) 242 Hypolepis 243 Paesia 244 Pteridium 245 H istiopteris 246 Lepidocau lon 247 Lonchitis 248 Blotiella

249 Saccoloma 250 Orthiopteris

Monachosoroideae 251 Monachosorum

lindsaeoideae 252 lindsaea 253 Humblotiella 254 Sambirania 255 Ormoloma 256 Sphenomeris 257 Odontosoria 298 Tapeinidium 259 Xyropteris

Thelypteridaceae 260 Thelypteris 26 1 Amauropelta 262 Phegopteris 263 Pseudophegopteris 264 Cyclogramma 265 Parathelypteris 266 Coryphopteris 267 Macrothelypteris 268 Oreopterrs 269 Metathelypteris 270 Cyclosorus 271 Trigonospora 272 Pronephr.ium 273 Mesophlebion 274 P tesioneuron 275 G laphyropteridopsis 276 Chingia 277 Hap lodictyum 278 Nannothelypteris 279 Stegnogramma 280 Steiropteris 28 1 Sphaerostephanos 282 Ampelopteris 283 Goniopteris 284 Meniscium 285 P neumatopteris 286 Christella 287 Amphineuron

Aspleniaceae: Asplenioideae 288 Asplenium 289 Camptosorus 290 Phyllitis 29 1 Ceterach 292 C eterach ops Is 293 Pieurosorus 294 Loxoscaphe 295 Die Ilia 296 Holodictyum 297 Antigramma 150 FERN GAZETTE : VOLUME 11 PART 2& 3 (1975)

Aspleniaceae: Asplenioideae (c ontinued) 298 Diplora 299 Schaffneria 300 Boniniella

A thyrioideae 301 Maneuccia 302 Onoclea 303 Onocleopsis

304 Athyrium 305 Cornopteris 306 Rhachidosorus

307 Diplazlum 308 M onomelanglum 309 Callipteris 310 Allantodia 311 D iplaziopsis 312 Dictyodroma 313 Anisocampium 314 Hemidictyum

315 Gymnocarp ium 316 Cystopteris 317 Acystopteris 318 Cystoathyrium

319 Woodsia 320 Lunathyrium 32 1 Dryoathyrium 322 A thyriopsis 323 Adenoderris 324 Cheilanthopsis

325 Hypodematium 326 Ku niwatsukia

Tectarioideae 327 Trichoneuron 328 Ctenitis 329 Ctenitopsis 330 Lastreopsis 331 Psomiocarpa 332 Atalopteris 333 Pleocnemia 334 Arcypteris 335 Pteridrys 336 Oryopolystichum

337 T ectaria 338 Luerssenia 339 Pseudotectaria 340 Hemigramma 341 Ouercifllix 342 Cionidium 343 Tectaridium 344 Fadyenia 345 P leu roderris 346 Hypoderris 347 Amphiblestra 348 Dictyoxiphium 349 Camptodlum 350 Stenosemia CRABBE, JERMY & MICKEL: A NEW GENERIC SEQUENCE 151

Asplen iaceae: Tectarioideae (con tinued) 351 Heterogonium 352 Cyclopeltis ------"-r- 353 Didymochlaena

Dryopteridoideae 354 Peranema 355 Diacalpe -- --- 356 Polystichum 357 Papuapteris 358 Plecosorus 359 Cyclodium 360 Cv,rtom ium 361 · Cyrtomidictyum 362 Cyrtogonellum 363 Phaneroph leb ia 364 Arachniodes 365 Polystichopsis 366 Lithostegia --·--- -- 367 Polybotrya 368 Maxonia ------369 Dryopteris 370 N othoperanema 371 Stigmatopteris ------372 Acrophorus 373 Stenolepia

Lomariopsidoldeae 374 Bolbitis 375 Eger10lfia 376 Thysanosoria 377 Arthrobotrya 378 Teratophylium 379 Lomagramm·a 380 Lomariopsis

E laphoglossoideae 38 1 E laphogiossum 382 Peltapteris 383 M icrostaphyla

Davalliaceae: Davallioideae 384 Humata 385 Trogostolon 386 Scyphularia 387 Parasorus 388 Davaliia 389 Davallodes 390 Paradavaliodes 39 1 Araiostegia 392 Leucostegia 393 Gymnogrammitis 394 Rumohra

Oleandroideae 395 Oleandra 396 Arthropteris 397 Psamm iosorus 398 Nephrolepis

Blechnaceae 399 Blechnum 400 Salp ichlaena 401 Doodia 152 FERN GAZETTE : VOLUME 11 PART 2 & 3 (1975)

Blechnaceae (c on tinued) 402 Brainea 403 Sadleria 404 Woodwardia 405 Anchistea 406 Chien iopteris 407 lorinseria 408 Pteridob lechnum 409 Stenoch laena

Marsileaceae 410 M arsilea 41 1 Regnellidium 412 Pilularia

Salviniaceae 413 Salvinia

Azollaceae 414 ' Azolla

ACKNOWLEDGEMENTS We should like to thank the following for commenting in part or in whole, on the arrangement. Dr Lenette R. Atkinson, Or D.W. Bierhorst, Prof. T.C. Chambers, Dr Barbara S. Croxall, G.J. De Joncheere, Or A. Murray Evans, Or E. Hennipman, Dr K. lwatsuki, Or Frances M. J'arrett, Or E.J. Klekowski, Or A. Lawalree, Or D.B. Lellinger, Or R.M. Lloyd, Prof. B.K. Nayar, Or C.N. Page, Prof. R.E.G. Pichi Sermolli, M.G. Price, Prof. E.A.C.L.E. Schelpe, Or A. Reid Smith, Or E.R. de la Sota, R.G. Stolze, Or Alice F. Tryon, Dr R.M. Tryon, Or W.R.J. Van Cotthem, Or W.H. Wagner, Jr., Dr T.G. Walker, Dr P. Zamora. Also we thank R. Ross and Or W.T. Stearn for help with nomenclature and Iatin diagnoses; and Miss Theresa Power for typing and preparing appendix 2 for lithography. lt is with particular pleasure that we record our gratitude to Dr R.E. Holttum whose almost every published work stresses phylogenetic relationships and stimulates pteridologists to continue to seek for more information and whose teaching in the herbarium is enlightening and enthusiastic.

REFERENCES BOWER, F.P. 1923-28 (reprint 1963: New Delhi). Th e ferns (Filicales). 3 vols. Cambridge. CHING, R.C. 1940. On natural classification of the family "Polypodiaceae ". Sunyatsenia 5: 201 -268. CHI NG, R.C. 1954. Systematic arrangement of families and genera of Chinese pteridophytes with corresponding names in C hinese. Acta phytorax. sin. 3: 93-99. CHRISTENSEN, C. 1906. E numeratio generum systematics cum synonym is omnibus chronologice et systematics ordinatis, in Index filicum: XI - Ll X. Copenhagen. CHRISTE NSEN, C. 1934. Enumeratio generum systematics hinc inde emendata, in Index filicum supplementum tertium: 5-15. Copenhagen. CHRISTE NSEN, C. 1938. Filicinae (pp 522-550) in F. Verdoorn (Ed.), Manual of pteridology. The Hague. COPE LAND, E .B. 1947. Genera filicum, the genera of fe rns. Waltham. DICKASON, F.G. 1946. The ferns of Burma. Cllio J. Sci. 46: 109-141 . ENGLER, A. & PRANTL, K. 1898-1902. Die naturlichen Pf/anzenfamilien 1 (4). Leipzig. HOLTTUM, R .E. 1947. A revised classification of leptosporangiate ferns. J. Linn. Soc. (Bot.) 53: 123-158. HOLTTUM, R .E. 1949. The classification of ferns. Bioi. Rev. 24: 267-296. HOL TTUM, R .E. 1966. I ntroduction to Pteridophyta [ + all pteridophyte entries] . In J.C. Willis, A dic tionary of the flowering plants and ferns, 7th ed. rev ised by H.K. Airy Shaw. Cambridge. HOL TTUM·, R.E. 1973a. Posing the problems. In A.C. Jermy et al. (Eds), The phylogen y and classification _ofthe fe rns. Bot. J. Linn. Soc. 67, Sup pi. 1 : 1-1 0. CRABBE, JERMY & MICKEL: A NEW GENERIC SEQUENCE 153

HO LTTUM, R.E. 1973b. Introduction to Pteridophyta [and) Family names of Pteridophyta [+ all pteridophyte entries) . In J.C. Willis, A dictionary of the flowering plants and fe rns, 8th ed. revised by H.K. Airy Shaw. Cambridge. JERMY, A.C., CRABBE, J.A. & THOMAS, B.A. (Eds). 1973. The phylogen yend classifica tion of the ferns. Bot. J. Linn. Soc. 67, Suppl. 1. MANTON, I. 1950. Problems of cytology and evolution in the Pteridophyta. Cambridge. MEHRA, P.N. 1961. Cytological evolution in ferns with particular reference to Himalayan forms. Proc. 48th lndian Sci. Congr. 2: 1-24. MICKE L, J.T. (Ed.) 1971. [Papers presented at the 1970 AIBS symposium entitled] Ferns as tools in solving biological problems. Bioscience 21: 225-227, 266-270, 271-275, 313-316, 317-322, 323-234. MICKE L, J.T. 1973. The classification and phylogenetic position of the Dennstaedtiaceae. In A.C. Jermy et al. (Eds), Th e phy/ogeny and classification of the ferns. Bot. J. Linn, Soc. 67, Suppl. 1: 135-144. MICKEL, J.T. 1974. Phyletic lines in the modern ferns. In T.N. Taylor & J.T. Mickel (Eds) , Evolution of sy.sterrfl tic characters in the fe rns. Ann. Mo. bot. Gdn 61 : 474-482. MORTON, C. V. 1968. The genera, subgenera and sections of the Hymenophyllaceae. Contrib. U.S. Net. Herb. 38: 153-214. NAYAR, B.K. 1970. A phylogenetic classification of the homosporous ferns. Ta xon 19: 229-235. PICHI SERMOLLI, R.E.G. 1958. The higher taxa of the Pteridophyta and their classification. In 0. Hedberg (Ed.), Systema tics of today. Proceedings of a symposium held at the University of Uppsala in commemoration of the 250th anniversary of the birth of Carolus Linnaeus. Uppsals Univ. J!.rsskrift 1958 (6): 7Q-90. PICHI SERMOLLI , R.E.G. 1973. Historical review of the higher classification of the Filicopsida. In A.C. Jermy et al. (Eds), Th e phylogen y and classification ofthe fe rns. Bot. J. Linn. Soc. 67, Suppl. 1 : 11-40. PRESL, K.B. 1836. Ten tamen pteridographiae. Prague. PRESL, K.B. 1845. Supplementum ten taminis pteridographiae. Prague. SMITH, J.E. 1793. Tentamen botanicum de filicum generibus dorsiferarum. Mem. Acad. r. Se. Turin 5: 401 -422. ' SMITH, J.E. 1810. Filices. In A. Rees. The cyclopaedia, or universal dic tionary of arts, sciences, and literature 14. l..ondon. SPORNE, K.R. (2nd ed. 1966). The morphology of pteridophytes. London. TAY LOR, T.N. & MICKE L, J.T. (Eds). 1974. Introduction [to papers (pp 310-482) presented at · the 1973 Amherst symposium entitled] Evolution of systematic characters in th e ferns. Ann. Mo. bot. Gdn 61: 307-309. TRYON, R.M. Jr. 1971 (ined.). Evolution and classification of the American tree ferns (Cyatheaceae). Presented at the 1971 AIBS symposium WA GNE R , W.H. Jr. 1969. The construction of a classification. In Systematic biology. U.S. Nat. Acad. Sci. Pub/. No. 1692: 67-90.

APPENDIX I Summary of the families and sub-families used in this paper Adiantaceae (C. Presl) Ching in Sunyatsenia 5: 229 (1940). Bas. Filicineae trib. Adiantaceae C. Presl, Te nt Pterid. 139 (1836 ). Subfam. Adiantoideae Subfam. Pteridoideae C. Chr. ex Crabbe, Jermy & Mickel subfam. nov. Syn.: Po/ypodiaceae subfam. Pteridoideae C. Chr. in Verdoorn, Man. Pterid. 536 (1938) pro parte; sine descr. lat. F ilices terrestres. rhizomate erecto vel repente plerumque solenostelico saepe perforate vel vere dictyostelico paleis induto. F rondes homomorphae tripartitaeque vel dimorphae pinnataeque, venis liberis vel anastomosantibus sine venulis inclusis. Sari marginales coenosori facientes, portati in commissuravas culare (qua extremitates venarum conjuncta suntl, margine folii membranaceo reflex a obtecti; vel sari faciens coenosorum super totam paginam sine receptacula extendentem vel in receptacula intramarginali inter venas positum. Sporae trlletae raro monoletae perispora praeditae. Chromosomata x = 29, 30. Typus: Pteris L, Sp. PI. 2: 1073 ( 1753). 154 FERN GAZETTE: VOLUME 11 PART 2 & 3 (1975)

Subfam. Vittarioideae (C. Presl) Crabbe, Jermy & Mic!

Aspleniaceae Mett. ex Frank in Leunis, Syn. Pflanze,nk., ed. 2, 3: 1465 (1877). Subfam. Asplenioideae. Subfam. Athyrioideae Holttum ex Nayar in Taxon 19: 235 (1970) sub Dryopteridaceae. Subfam. Dryopteridoideae Holttum ex Nayar in Ta xon 19: 235 ( 1970) pro parte, sub Dryop teridaceae. Subfam. Elaphoglossoideae (Pichi Serm.) Crabbe, Jermy & Mickel stat. nov. Basionym: Elaphoglossaceae Pichi Serm. in Webbia 23: 209 (1968). Syn.: Po/ypodiaceae subfam. E/aphog/ossoideae C. Chr. in Verdoorn, Man. Pterid. : 549 (1938), sine descr. lat. Subfam. Lomariopsidoideae (Aiston) Crabbe, Jermy & Mickel stat. nov. Basionym: Lomariop sidaceae Alston in Tax on 5: 25 (1956), excl. Elaphog/ossum. Syn.: Dennstaedtiaceae subfam. Lomariopsidoideae Holttum in J. Linn Soc. Lond., Bot, 53: 146 (1947), sine descr. lat., excl. E/aphog/ossum. Subfam. Tectarioideae Holttum ex Nayar in Tax on 19: 235 (1970) sub Dryop teridaceae.

Azollaceae Wettstein, Hmdb. Syst. Bot. 2: 77 (1903).

Blechnaceae (C. Presl) Copeland, Gm. Fil. 155 (1947).

Cheiropleuriaceae Nakai, in Bot. Mag. , Tokyo 42: 210 (1928).

Cyatheaceae Kaulfuss, Wesen Farrenkr. 119 (1827).

Davalliaceae Mett. ex Frank in Leunis, Syn. Pflanzenk., ed. 2, 3: 1474 (1877). Subfam. Davallioideae. Subfam. Oleandroideae (Ching ex Pichi Serm.) Crabbe, Jermy & Mickel, stat. nov. Basionym: 0/eandraceae Ching ex Pichi Serm. in Webbia 20: 745 (1965). Syn.: Po/ypodiaceae subfam. 0/eandroideae C. Chr. in Verdoorn, Man. Pterid. 365 (1938), sine descr. lat.; 0/eandraceae Ching in Sunyatsenia 5: 221 (1940), sine descr. lat.; Dennstaedtiaceae· subfam. Oleandroideae Holttum in J. Linn. Soc. Lond., Bot. 53: 137 (1947), sine descr. lat.; Davalliaceae subfam. 0/eandroideae (C. Chr.) Tindale in Con tr. N.S. Wa les Nat. Herb., Fl. Ser. 208: 7 (1961), comb. illeg. CRAB BE, JERMV & MICKEL: A NEW GENERIC SEQUENCE 155

Dennstaedtiaceae Pichi Serm. in We bbia 24: 704 (1970). Subfam. Dennstaedtioideae Subfam. Lindsaeoideae (Pichi Serm.) Crabbe, Jermy & M ickel, stat. nov. Basionym: Lindsaeaceae Pichi Serm. in Webbia 24 : 707 (1970). Syn.: Po/ypodiaceae subfam. Lindsayoideae C. Chr. in Verdoorn, Man. Pterid: 535 (1938), sine descr. lat.;Dennstaedtiaceae subfam. Lindsayoideae Holttum in J. Linn. Soc. Land., Bot. 53 : 134 (1947), sine descr. lat. Subfam. Monachosoroideae Crabb�, Jermy & Mickel, subfam. nov. F ilices terrestres, rhizomate ascendenti vel erecto dictyoste1ico, sine paleis sed frondibus caulibusque juvenilibus muco et pUis minutissimis indutis. Frondes pinnatae ad quadripinnatae, homomorphae, non articulatae: stipes mucilaginus hirsutus, duobus fasciculis vasorum ut in Athyrio; lamina tenuis in statu vivo fare pellucida sad in siccitate impellucida, venis liberis. Sori superficiales, ad ap ices ex venarum te rminales saepe sporangiis paucis constantes, exindusiatl. Spo· rae triletae perispora praeditae. Chromosomatax = 56. Typus: Monachosorum Kunze in Bot. Zeit. 6: 119 ( 1848) . Syn.: Monach osoraceae Ching in Sunyatsenia 5: 241 ( 1940) , sine descr. lat.

Dipteridaceae (Diels in Engler & Prantl) Seward & Dale in Phi/os. Trans. Roy. Soc. Lond ser. B. 194: 487, 499, 502 (1901 ). Bas. Po/ypodiaceae trib. Aspidieae subtrib. Dipteridinae Diels in Engler & Prantl, Nat. Pf/anzenfam. 1(4): 167, 202 (1899).

Equisetaceae Rich. ex DC. in Lam. & D.C., Fl. Franc. ed. 3. 2: 580 (1805). Gleicheniaceae (R. Br.) C. Presl, Ref. Haenk. 1(1): 70 (1825). Grammitidaceae (C. Presl ) Ching, Sunyatsenia 5: 264 (1940) ("Grammitaceae"). Bas. Filicaceae trib. Grammitaceae Presl, Tent.Pt erid. 205 (1836). Hymenophyllaceae Link, Handb. Erken. Gew. 3: 36 (1833) ("Hymenophylleae"). Hymenophyllopsidaceae Pichi Serm. in Webbia 24: 712 (1970). lsoetaceae Dumontier, Ana/. Fam. PI. 67 (1829) ("lsoetineae"). Lophosoriaceae Pichi Serm. in Web bia 24: 700 (1970). Loxsomaceae C. Presl, Gefassb. Stip . Farm. 31 (1847). Lycopodiaceae P. Beauv. ex Mirbel in Lamarck & Mirbel, Hist. Na t. Veg. 4: 293 (1802 ) ("Lycopodia"). Marattiaceae Berchtold & J.S. Presl, Prirozen. Rost/. 1: 272 ( 1 �20) ("Marattiae"). Marsileaceae Mirbel in Lamarck & Mirbel, Hist. Nat. Veg. 5: 126 (1802 ) ("Marsileae"). Matoniaceae C. Presl, Gefassb. Stip. Farm 32 (1847). Metaxyaceae Pichi Serm. in Webbia 24: 701 (1970). Ophioglossaceae (R. Br.) Agardh, Aphor. Bot. 8: 113 (1822). Bas. Filices trib. Ophioglosseae R. Br., Prodr. F/. Nov. Ho/1. 163 (1810). Osmundaceae Berchtold & J.S. Presl, Prirozen Rostl. 1: 272 (1820). Parkeriaceae Hooker, Ex ot. Fl. 2_(20): t. 147 (1825). Plagiogyriaceae Bower, Ann. Bot. 40: 484 ( 1926). Platyzomataceae Nakai, Bull. natn. Sci. Mus. To kyo 29: 4 (1950). Polypodiaceae Berchtold & J.S. Presl, Prirozen Rostl. 1: 272 (1820). Subfam. Polypodioideae 156 FERN GAZETTE: VOLUME, 11 PART 2 & 3 (1975i

Subfam. Drynarioideae Crabbe, Jermy & Mickel, sub�am. nov. Filices epiphyticae; rhizome saepe crassum, carnosum, lepldotum, raro tenue glabrescensque. Frondes glanduliferae, cum rhizomate non vera articulataf!, saepe eimorphae vel basibus frondium scari05is per humi cotlectionem praesertim mutatis (quamobram frondes nidi vel cornucopias feciuntl; pinnae vel segmenta fron�lium normalium (i.e. frondium foliacearum) semper ab rhachide caducae, texture coriacee herbac&ll\!8, venis conspicuis retlculatis areolatis venulis lnclusis. Sorl ad compitum venarum punqtati vel coenosori facientes; perphyses desunt . Spciree monoletae;pler umque perisporaedesu nt Chromosomata x 36, 37. Typus: Drynaiia (Bory, pro 'sectione Polypodii in Ann. Sci. nat 5: 464) J. Smith in J. Bot Lond. 4: 60 (1841). Subfam. Microsorioideae Nayar in Taxon 19: 233 (1970). Subfam. Platycerioideae Nayar in Taxon 19: 233 (1970). Subfam. Pleopeltoideae Nayar in Tax on 19: 234 (19_70).

Psilotaceae Kanitz, Novenyrends. Attek. 43 (1887). Salviniaceae Dumortier, Anal. Fam. PI. 67 (1829). Schizaeaceae Kaulfuss, We sen Farrenkr. 119 (1829). Selaginellaceae Willkomm in Willkomm & Lange, Prodr. Fl. Hisp. 1 (1): 14 ( 1861 ). Stromatopteridaceae (Nakai) Bierhorst in Phytomorphology 18: 263 ( 1968). Bas. Gleicheniaceae subfam. Stromatopteridoideae Nakai in Bull. natn. Sci. Mus. To kyo 29: 32 (1950). Thelypteridaceae Ching ex Pichi Serm. in Webbia: 709 (1970). Thyrsopteridaceae C. Presl, Gefassb. Stip. Farm, 38 (1847 ) [Abh. K. Bohm. Ges. Wiss., ser 5, 5: 346 (1848)] . APPENDIX 2

Generic nameS. This list is based on Holttum ( 1966, 1973b) and In the majority of cases we have accepted the concepts of Holttum with regard to the synonymy. Some names of minimal importance (e.g. orthographic variants) have not been included because this list is primari ly a working list for herbarium curators. Those genera asterisked are nomina conservanda.

Ab acopteris Fee = 272 Alsophila R.Br. 2j1 Ab rodictyum Presl 122 Amauropelta Kunze 261 Aconiopteris Presl = j81 Amblia Presl = 363 Acropelta Nakai = 356 Amesium Newm . = 288 Acrophorus Presl 372 Ampelopteri s Kunze 282 Acropteris Link = 288 Amph iblestra Presl 347 Acropt erygium (Diels )Nakai 1�8 Amphicosmia Gardn . = 23 1 Acrorumohra (H.Ito)H.Ito 369 Amphidesmium J.Smith = 225 Acrosorus Copel . 21 4 Amphineuron Holtt. 287 Acrostichum L. 101 Amphipterum Presl = 115 Actini opt eris Link 039 Amphisoria Trev . = 331 Actinophleb ia Presl = 227 Amphoradenium Desv . 21 5 Actinostachys Wall . 03j Anantha corus Underw . & Maxon 092 Acystopteri s Nakai 31 8 Anapausia Presl 202 Adectum Link = 2j8 Anapeltis J.Smith = 186 Adenoderris J.Smith 323 Anarthropteris Copel . 224 Adenophorus Gaud . = 215 Anaxetum Schott = 200 Adiantopsis Fee 048· Anchistea Presl 405 · Adiantum L. 087 Aneimia Sw . corr .Kaulf. = 035 Aetopteron Ehrh . = 35b Aneimiaebotrys Fee = 035 Afropteris Alston 040 Aneimidictyon J.Smith corr .C .Chr.= 034 · Aglaomorpha Schott 156 *Anemia Sw . 035 Alcicornium Gaud . = 1 b2. Anemidictyon J.�mith = 035 Aleuritopteris Fee 049 Anemirhiza J,Smith = 035 Allantodia R.Br . ( 1810) = 307 Anetium (Kunze)Splitg. 091 Allantodia Wall . ( 1850 ) = j11 Angiopteris Adans . ( 176j ) '= j02 Al losorus Bernh . = 047 *Angiopt eris Hoffm . ( 1796) 019 Allosorus auctt . = 044 Ani socampium Presl 313 Allost el:it es BC:lrner = 01 1 Ani s ogonium Presl 307 CRABBE, JERMY & MICKE L: A NEW GENERIC SEQUENCE 157

Ani sosorus Trev .(nomen) = 247 Campyl ogramma v.A.v.R. = 166 Anogramma LinK 062 Candollea Mirbel = 163 Anopt eris ( Prantl )Di.els 042 Cardiochlaena Fee = 337 Ant igramma Presl 297 . Cardiomanes Presl 116 Ant iosorus Roemer 247 Carpanthus Rafin . = 41 4 Ant rophyum Kaulf. 087 Carpogymnia Love & Love = 315 Apalophlebia Presl = 163 Cassebeera Kaulf. = 059 Aphyllocalpa Cav . 026 Cephaloceraton Genn . = 009 Aptercpteris ( Copel . )Copel . 112 Cephalomanes Presl 127 Arachniodes Blume 363 Ceratodactylis J.Smith = 045 *Araiost egia Copel . 391 Ceratopteris A.Brungn . 037 Archangiopt eris Chri st & Ceropteris Link = Ob3 Gi esenhagen 020 Cerosora Bak . 085 Arcypteris Underw. 334 Ceterac Adans . = 288 Argyrochosma (J.Smlth)Pic .Ser. 054 *Ceterach Lam . & DC . 29 1 Aropteris Alston (sphalm .) = 040 Ceterachopsis (J.Smith)Ching 292 Arthrobotrya J.Smith 377 Chamaeclinis Mart .(nomen ) = 003 Arthrobotrys Wall .(nomen) = 369 Chamaefilix Hill = 292 Arthromeris (Moore )J .Smith 183 *Cheilanthes Sw. 047 Arthropteris J.Smith 306 Cheilanthopsis Hieron. 324 Aspidotis (Nutt.)Copel . 050 Cheilogramma ( Bl . )Maxon = 191 Aspidium Sw . = 337 Cheilolepton Fee = 379 Asplenicystopteris P.Fourn . 31 6 Cheiloplecton Fee 051 Asplenidictyum J.Smith = 288 Cheilosoria Trev. = 047 · Aspleniopsis Mett. 083 Chei roglossa Presl 01 7 Asplenium L. 288 Ch ei ropleuria Presl 151 Asplenoceterach D.E .Meyer 288 x 291 Chei ropteris Christ = 198 Asplenophyllitis Alston 288 x 290 Chi eniopt eris Ching 406 Asplenosorus Wh erry 288 x 289 Chingia Holtt . 276 Asteroglossum J.Smith (nomen) = 165 Chnoophora Kaulf. = 229 Atalopt eris Maxon & C.Chr . 332 Chori stosoria Mett . = 057 Athyriopsis Ching 322 Chorizopteris Moore = 379 Ath,yrium Roth 304 Christella Lev.emend .Holtt . 286 Austrogramr.1e ,F ourn . 078 Christensenia Maxon 025 Azolla Lam . 41 4 Chri stiopteris Copel . 1J6 Chrysochosma (J.Smith )Kilmm . 049 Bakeriopteris O.Ktze = 058 Chrysodium Fee = 101 Balant ium J:aul f. = 233 Chrysopteris Link = 204 Balant ium auctt. = 236 Cibot ium Kaulfuss 237 Bathmium Link = 337 Cincinalis Gled . ( 1764 ) = 047 Belvisia Mirbel 196 Cincinalis Desv . ( 1811) = 244 Bernhardia Willd . = 001 Cionidium Moore 342 Blechn�dium Moore = 399· Clementea Cav . = 019 Blechnopsis Presl = 399 Clopodium Rafin .(nomen ) 003 Blechnopteris ·Trev .( nomen) 399 Cnemidaria Presl 227 Blechnum L. 399 Cnemidopteris Rei chenb . 227 Blotiella Tryon 248 Cochlidium Kaulfuss 221 Bolbitis Schott 374 Colina E.L.Greene = 036 Bommeria Fourn . 068 Colysis Presl 170 Boniniella Hayata 300 Coniodictyograrr,me Nakai = 082 Boschia Copel . (nom .prov) = · 115? *Coniogramme Fee 082 Botrychium Sw . 01 2 Copelandi opt eris B.C.Stone 098 Botryogramme Fee = 045 Copodium Rafin. = 003 Bot ryopteris Presl = 015 Coptidipteris Nakai & Momose 252? Botrypus Ri.chard 013 Coptophyllum Gardner = 035 Bowringia Hooker = 402 Cormophyllum Newm . 228 Brachysorus Presl = 304 Cornopteris Nakai 305 Brainea J.Smith 402 Coryphopteris Holtt . 266 Buesia (Mort on)Gopel . 107 Cosentinia Todaro = 047 Byrsopteris Morton = 364 Costaricia Chri st = 238? Craspedaria Link 188 Caenopteris Bergius = 288 Craspedodictyum Copel . 076 Calamaria Boehm . = 009 Craspedoneuron v.d.Bosch = 123 Calamistrum O.Ktze = 41 2 Craspedophyllum ( Presl· )Copel . 113 Callipteris Bory 309 Crepidium Presl = 121 Callistopteris Copel . 128 Crepidomanes (Presl )Presl (1851 ) 119 Callogramme Fee = 075 Crepidomanes v.d.Bosch (1861 ) = 121 Calymella Presl = 144 Crepidophyllum Reed = 121 Calymmodon Presl 21 3 Crepidopteris Copel . = 121 Calypte rium Bernh . = 302 Crypsinus Presl 181 Campium Presl = 374 Cryptogramma R.Br . 044 Campt eria Presl = 096 . cryptosorus Fee = 21 2 Camptodium Fee 349 Cteisium Mi chx = 034 Camptosorus Link 289 Ctenitis (C.Chr .)C .Chr. 328 Campyloneurum Presl 201 . Ctenitopsis Ching 329 ' 158 FERN GAZETTE·: VOLUME 11 PART 2 & 3 (1975!

Ctenopteris Blume 21 2 Dryopolystichum Copel . 336 Culcita Presl 236 *Dryopteri s Adans . 369 Currania Copel . = 31 5 Dryostachyum J.Smith 157 Cuspidaria Fee = 197 Cyathea Smith 22b Eatoniopt eris Bon.mer = 232 Cyclodium Presl 359 Edanyoa Copel . = 375 ? Cyclograrnma Tagawa 264 Egenolfia Schott 375 Cyclopeltis J .Sru ith 352 *Elaphoglossum Schott )81 Cyclophorus Desv . = 163 Ellebocarpus Kaulf . = 037 Cyclopt eris Gray = )1 6 Enierosora Bak . = 21 7 Cyclosorus Link 270 Equisetum L. 01 1 Crytogonellum Ching 362 Eremopodium Trev . 288 Cryt ogonium J.Smith = 374 Erio�orus Fee 073 Cyrt omidictyum Ching 361 Eschatogramme Trev. = 197 Cyrt omium Presl 360 Eup odium J.Smith = 022 Cy rtophlebium (R.Br. )J .Smith 201 Eup.t eris Newm . = 244 Cystea Smith = 31 6 Cystoathyrium Ching 318 Fadyenia Hook , 344 Cy stodium J.Smith 234 Feea Bory 139 *Cystopt eris Bernh . 31 6 Filicula Segui er = 31 6 Filinguis Rafin . = 290 *Danaea Smith 024 Filix Segui er = 369 Danaeopsis Presl (gen .dub.) Fourniera Bommer = 232 Darea Juss . = 288 Furcaria Desv . = 037 Davallia Smith )88 Fuziifilix Nakai & Momose 239 Davalliopsis v.d.Bosch 126 Daval lodes (Copel . )Copel . 389 Galeoglossa Presl = 163 Dendroconche Copel . 167 Gisopteris Bernh . = 034 Dendroglossa Presl 175 · Glaphyropteridopsis Ching 275 Dennstaedt ia Bernh . 238 Glaphyropt eris Presl = 260 Deparia Hook . & Grev . 304 Gl eichenella Ching = 148 Dermatophlebium Presl = 111 *Gleichenia Smith 144 Diacalpe Blume 355 Gleicheniastrum Presl = 144 Diblemma J.Smith 172 Gl ossopt eris Rafin , = 290 Dichasium (A.Braun )Fee = 369 Glyphotaenium J,Smith 21 7 Dichorexia Presl = 2)1 Goniophlebium (Bl,)Presl 207 Dicksonia L'Herit . 235 Goniopt eris Presl 283 Diclidopteris Brack . = 095 Gonocormus v.d.Bosch 124 'Diclisodon Moore = 369 Grammatopteridium v.A.v.R. 178 Dicranoglossum J.Smith 197 Grammatopteris v.A.v.R. = 178 Dicranopt eri s Bernh . 147 Grammiti s Sw . 210 Dictymia J,Smith 203 Grammat osorus Regel = 337 Dictyocline Moore = 279 Guerinia J.Smith = 252 Dictyodroma Ching 31 2 Gyrnnocarpium Newm . 31 5 Dictyoglossum J.Smith = 381 Gymnogran.ma Desv. = 067 Di ctyogramme Fee = 082 Gyrnnogramme auctt . = 073 Dictyopt eris Presl = 334 G:vmnogrammitis Griffith 393 Dictyoxiphium Hook.348 Gymnogynum P.Beauv . = 008 Didiclis P.Beauv . = 008 Gymnopteris Bernh . 067 Didymochlaena Desv . 353 Gymnosphaera Bl . 231 Didymoglossum Desv.( 1827) 130 Gymnotheca Presl 022 Didymoglossum Prantl (1875 ) = 119 Gy rosorium Presl 163 Diellia Brack . 295 Digrammaria Presl = 307 Habrodictyon v.d.Bosch 122 Dimorphopteris Tagawa & Iwats . 270 Haplodictyum Presl 277 Diphasium Presl 005 Haplopteris Presl = 093 Diplaziopsis C.Chr . )1 1 Hecistopteris J.Smith 091 Diplazium Sw , 307 Helminthostachys Kaulf. 005 Diploblechnum Hayata = 399 Hemesteum Leveille = 356 • Diploophyllum v.d.Bosch = 115 Hemestheum Newm . = 260 Diplopt erygium (Diels)Nakai 145 Hemianemia (Prantl )Reed = 035 Diplora Bak . 298 Hemicardi on Fee = 352 Diplostachyum P.Beauv . = 008 Hemicyatheon Domin & Copel . 114 Dipteris Reinw . 152 Hemidictyum Presl , 31 4 Discostegia Presl = 022 · Hemigramma Christ 340 Di sphenia Presl = 228 Hemionitis L. 066 Distaxia Presl = 399 Hemiphlebium Presl = 130? Doodia R.Br . 401 Hemipteris Rosenst . = 096 Dorcapteris Presl = 367 Hemistachyum Copel . = 156 Doryopteris J.Smith 059 Hemistegia Presl = 227 *Drymoglossum Presl 165 Hemitelia R.Br. = 228 Drymotaenium Makino 194 Heterodanaea Presl = 024 Dryna ria (Bory )J .Smith 153 He�erogonium Presl 351 Drynariopsis (Copel . )Ching 159 Heteroneuron Fee = 374 Dryoathyrium Ching 321 Heterophlebium Fee = 096 Dryomenis Fee = 337 Heterophyllium Hieron . = 008 CRABBE, JERMY & MICKEL: A NEW GENERIC SEQUENCE 159

Heteropteris Fee ( 1843 ) = 191 Leptolepia Mett . 241 Heteropt eris Fee (1869 ) = 058 Lept ophyllurn Rafin . = 062 Hewardia J.Smith = 086 Leptopieuria Presl = 39 8 Hi criopt eris Presl (1849 ) = 147 Lept opteris Presl 030 Hicriopteris sensu Copel . ( 1941 ) 1 45 Lept orumohra H.Ito = 369 ? Leptostegia D.Don = 043 Hippochaete Mi lde = 01 1 Leucornanes Presl = 123 Histiopteris (Ag. ) J.Srnith 245 Leucostegia Presl 392 Hoffmannia Willd. = 001 Holcosorus Moore 180 Llndsae"a Dryand . 252 . Holodictyum Maxon 296 Lindsaya Dryand .(Kaulf. ,1b24; sphalm . ) ' = 252 Holostachyum (Copel . )Ching 158 Lindsaynium Fee = 254 Holttumia Copel . = 079 Lindsayopsis Kuhn = 257 Holtturniella Copel . 081 Li thostegia Ching J66 . Hornalosche Ehrh . = 005 Litobrochia Presl = 096 Homalosorus Small = 304 Llavea Laga s ca 045 Homoeotes Pre sl 138 Lon;agramma J .Smith 379 Homophyllum Merino = 399 Lomaphlebia J .Smith = 21 0 Humata Cav . 384 Loma ria Willd . = 399 Hurnblotiella Tard . 253 Lomaridium Presl = 399 Hup erzia Bernh . 006 Lomariobotrys Fee = 409 Hyalolepi s Kunze = 196 Lomariopsis Fee 380 Hyalotricha Copel . 222 Lonchitis L. 247 Hydroglossum Willd. = 034 Lophidium Rich. = 032 Hymenasplenium Hayata = 288 Lophodium Newm . = 369 Hym enocystis C.A .Meyer = 3 19 LopholeQ is J .Smith = ·1 86 Hyrnenodium Fee = 381 Lophosoria Presl 226 Hymenoglossurn Presl 103 Lorinseria Presl 407 Hymenolepis Kaul f. = 196 Lotzea Kl otz . & Ka rst. = 307 Hyrnenophyllopsis Goebel 142 Loxogramrne (Bl.)Presl 223 Hyrnenophyllurn Smith 106 Loxoscaphe Moore 294 Hyrnenostachi s Bory = 139 Loxsoma R. Br . 101 Hypochlamys Fee = 304 Loxsomopsis Chri st 102 Hypodematium Kun�e 325 Luerssenia Kuhn 338 Hypoderris R.Br . 346 Lunaria Hill = 01 2 Hypolepis Bernh . 242 Luna thyrium Koidsu11ci 320 Hypopeltis Michx = 356 Lycopodiella Holub 004 Hypopterygiopsis Sakurai 008 Lycopodina Bub . & Penz , = 008 Lycopodi oides Boehm . = 008 Idiopteris T.G. Walker 097 Lycopodium L. 003 Isoetes L. 009 Lygodi ctyon J.Smith = 034 Isoetella Genn. = 009 *Lygodium Sw . 034 Isoloma J.Smith (1841 ) = 252 Isoloma J.Smith (1842 ;1875 ) = 398 Macroglena ( Presl )Copel . 125 Ithycaulon Copel . = 250 Macroglossum Copel . 023 Macroplethus Presl = 196 Jamesonia Hook . & Grev . 072 Macrostowa Griff . = 025 Japanobotrychium Masamune 013 Macrothelypteris Ching 267 •.= Jenkinsia Hook 374 Marattia Sw . 022 Marginaria Bory 209 Kaulfussia Bl . = 025 Marginari opsis C.Chr. 190 Kaulinia Nayar = 166 Marsiglia Rafin. = 41 0 Ku niwat sukia Pic.Ser . 326 Marsilea L. 41 0 Marsilla Rafin. = 41 0 Lacaussadea Gaud . = 375 Matonia R.Br. 149 cos e v.d.Bosch La t a 136 *Matteuccia Tod . 301 Lastrea Bory s.l. · = 260 l.laxonia C.Chr. ·368 Lastrella (H.Ito)Nakai = 262 Mecodium Presl 115 reop Cning Last sis 330 Mecosorus Kl ot z. = 210 Lathyropteris Christ = 096 Menopt eris Rafin . (nornen) = 01 2 Lecanolepis Pic.Ser . 132 Meni scium Schreb . 284 Lecanium Presl = 1)2 Menisorus Al ston = 271 '/ ; 284? Lecanopt e ris Reinw . 169 Meringium Presl 108 Lemma Adans . = 41 0 Mertensia Willd . = 146 Lemmaphyllum Presl 193 Merinthosorus Copel . 155 Lewapteris Rafin . = 096 Mesochlaena R.Br. = 281 Lenda Koidz . = 337 Mesoneuron Ching = 273 Lepicystis J.Smith = 206 Mesophlebi on Holt t. 273 Lepiaocaulon Copel . 246 Mesosorus Hassk . = 146 Lepidogrammitis Ching = 1 93 Me s othema Presl = 399 Lepidoneuron Fee = 398 Metathelypt eris Ching 269 Lepidotis P.Beauv . = 004 l.letaxya Presl 225 Lepisorus (J.Smith)Ching = 185 Microbrachia Presl .= 337 Leptochilus Kaul f. 173 Microchlaena Ching = 326 Lept ocionium Presl = 111 Microgonium Presl 131 Lept ogramma J.Smith = 279 Mi cr ograrnma Presl 186 160 FERN GAZETTE:VO LUME 11 PART 2 & 3 (1975)"

Microlepia Presl 239 Olfersia Raddi = 3 67 Micropodium Mett. = 298 Oligocampia Trev . = 51 3 Micropolypodium Hayata = 21 1 Onoclea L. j02 Micropteris Desv . = 21 1 Onocleopsis F.Ballard 303 Microschizaea Reed = 032 Onychium Kaulr.( 1 820 ) 043 Microsorum Link 166 Onychium Reinw .( 1825 ) 169 Microsorium auctt.( sphalm .) 166 Oochlamys Fee = 261 Microstaphyla Presl 383 Ophiala Desv . � 01 5 Microstegia Presl = 307 Ophioderma (Bl.)Endl . = 01 6 Microstegnus Presl = 227 Ophi ogl ossum L. 01 6 Microterus Presl = 181 Ophiopteri s Reinw . = 395 Microtrichomanes (Mett. )Gopel . Oreogramn.i tis Copel . 21 8 ( gen . dub . ) 1 Ob s .l • ; 117 s .1. Oreo�t eri s Holub 2bb Ormoloma Maxon 255 Mildella Trev . ·052 Mirmau Adans . = oo8 Orrnopteris J.Smith 059 Mohria Sw . 03b Ornithopteris Bernh . = 0)5 Monachosorella HByata 251 Ortniopteris Copel . 25U Monachosorum Kunze 251 Orthogramma Presl = 599 Monochlaena Gaud . = 353 Osmunda L. 026 Monogonia Presl = 283 Osmundastrum Presl 027 Monogran•ma Schkuhr 094 Osmundopteris (Milde )Small 013 Monomelangium Hayata 308 Oxgonium Presl = 307 Myriodon Copel . 110 Myri opt eris Fee = 0�7 Pachyloma v.d.Bosch = 113 Myriotheca Cornu, . ex JusG . = 0 22 Pachypleuria Presl = )84 = Myrrnecophila (Chri st )NaKui 169 Paesia St .Hil . 243 Myrmecostylum Presl = 108 Palhinhaea Franco & Vasconc . 004 Myuro�teris C.Chr. = 175 Palt onium Presl = 191 Panicularia Colla = 235 Nanothelypt eris Holtt . 278 Papuapteris C.Chr . 357 Negripteris Pic.Ser . 056 Parablechnum Presl = 399 = Nen.at opera Kunze 354 Paraceterach( F.Muell .)Copel . 069 Nernatopt eris v.A.v.R. 219 Paradava llodes Ching 390 Neoch eiropteris Christ 198 Paradennstaedtia Tagawa = 238 Neolepisorus Ching 192 Paragral!,ma ( Bl . )Moore 195 = Neoniphopsis Nakai 103 Paraleptochilus Copel . 174 Neottopteris J.Slliith = 28o Parasorus v.A.v.R. 387 Nephelia Tryon 230 Parathelypteris Ching emena ,Holtt . 265 Nephopteris Lell1nger 074 Parathyrlum Holtt . = 321 Nephrodium Rich .( gen ,dub.)= 260; 369 Parestia Presl 388 Nephrolepis Schott 398 Parkeria Hook . = 037 Nesopteris Copel . 129 Patania Presl = 238 Nesosoris Rarin. = 1b3 *Pellaea Link 057 Neurocallis Fee 099 Pellaeopsi s J.Srr.1 th = 05 7 Neurodium Fee 191 Peltapteris Link 382 Neurograthn.a Link = 067 Peltochlaena Fee 371 Neuromanes Trev . 134 Pentarn1zidium Hayata = 301 Neuronia D.Don = 395 Peranema D.lJon 354 Neurophyllum Presl = 134 Peripteris Rarin. = 096 �europlatyceros (Endl .)Fee = 162 Pessopteris Unaerw . & Maxon 2u0 Neuropteris Gaud , = 249 Phanerophleb ia Presl 363 Neurosoria Mett . 046 Phanerophleb iopsis Ch1ng = 356 Neurosorus Trev .(nomen) CJ82 Phanerosorus Copel . 150 Niphidiurn J.Smith 199 Phegopteris (Presl )Fee emend . Niphobolus Ka ulr. = 163 Ching 262 Niphopsis J.Smith = 163 Phl eb iogonium Fee = 337 Notogramme Presl = 082 Phl ebi ophyllum v.d.Bosch = 120 Notolep eurn Newm . = 291 Pnlebodium (R.Br. )J .Smith 204 Notholaena R.Br. 053 Phl egmariurus Holub = 006 Nothochlaena auctt . = 053 Phorolobus Desv . = 044 Nothoperanema (Tagawa )Ching 371 Photinopteris J.Smith 154 Phyllitis Ludw . ( 1757) 290 Ochlogramn.a rpesl = 307 Phyllitis Rarin . (1819) = 096 Ochropteri.s J .Smith 041 Phylloglossum Kunze 007 Odont olon.a J .Srr.ith = 252 Phymatodes Presl = 152 Odontomanes Presl 135 Phymatop sis J.Smith = 181 Odonto1-teris Bernh . = 034 Phymatopteris Pic .Ser. = 181 Odont osoria Fee 257 Phymat osorus Pic .Ser. 168 Oenotrichia Copel . 240 . Physematium Kau�f . = J1 9 Oet osis O.Ktze .(nom .dub.) 079 ;093 ; Pilularia L. 41 2 165; 191 Pinonia Gaud . = 237 Pityrogramma Link 063

Oet osis Greene = 093 Plagiogyria (Kunze)Mett . 031 Oleandra Cav . 395 Plananthus P.Beauv . = 006 Oleandropsis Copel , 179 Platycerium Desv . 162 CRABBI;·, JERMV & MICKEL: A NEW GENERIC SEQUENCE 161

Platyloma J.Smith � 057 Quercirilix Copel . 341 Platytaenia Kuhn 080 Platyzoma R.Br . 038 Ragatelus Presl 140 Plecosorus Fee 358 Ragi opteris Presl � 302 Plect opteris Fee � 21 3 Rarnondia Mirbel � 034 · Plenasium Presl 028 Reedi ella Pic.Ser . 121 Pl eocnemia Presl 333 Regnellidiurn Lindrnan 41 1 Pl eopeltis Hurnb . & Bonpl . 185 Rhachidosoru,s Ching 306 Plesioneuron (Holtt .)Holtt . 274 Rheopteris Al ston 084 Pleuridium (Presl) Fee � 200 Rhipidopteris Schott � 382 Pleuroderri s Maxon 344 Rhizoglossurn Presl 01 8 Pleurorossa Nakai � 094 Rhizosperrna Meyen � 41 4 Pleurograrnme (Bl.) Presl 094 Riedlea Mirbel � 302 Pleuromanes Presl 123 Ripidium Bernh . � 0)2 Pleurosoriopsis Fornin 065 Rosenstockia Copel . 105 Pleurosorus Fee 293 Rurnonra Raddi 394 Pneumatopteris Nakai 285 Rumohra auctt . � 356 Podopeltis Fee � 337 Rutarnuraria Ort . � 288 Podosorus Holtt . 175 Poecilopteris Eschw . � 374 Saccolorna Kaulr. 249 Polybot rya· Huu,b . & Bonpl . 367 Sadleria Kaulr. 40 1 Polycarnpiurn Presl � 163 Sarrordia Maxon 060 Polycocca Hill � 088 Sagenia Presl �·337 Polydictyurn Presl � 337 Sal pi chlaena J.Smith' 400 Polyphlebium Copel . 120 Sarnbirani a Tard . 254 Polypodiopsis Copel . � 184 Salvinia- Seguier 41 3 Polypodiopt eri s Reed 184 Saxiglossum Ching 164 Polypodiurn L. 206 Sceptridium Lyon 01 4 Polystichopsis (J.Srr,ith )Holtt . 365 Schaffneria Fee 299 Polystichurn Roth 356 Schellolepis J.Srn1th � 207 Polytaeniurn Desv . 088 *Sch izaea Smith 032 Pornataphyturn Jones � 047 Schizocaena J.Smith � 232 Prorerea Presl � 270? Schizogramma Link (gen .dub.) 066? Pronephriurn Presl 272 Schizol egnia Al ston � 252 Prosapt ia Presl 21 6 Schizolepton Fee � 079 Protangiopteris Hayata � 020 Schi zolorna Gaud . � 252 · Protol indsaya Copel . � 258 Schi zoloma sensu Copel . � 079 Protornarattia Hayata 021 Schizostege Hilleb. � 096 Protowoodsia Ching � 319 Schi zostegeopsis Copel .( sphalm. ) 098 Psammiosorus C.Chr. 397 Schi zostegopsis Copel . � 098 Pseudathyriurn Newm . � 304 Scleroglossum v.A.v.R. 220 Pseudocyclosorus Ching � 285 Scoliosorus Moore 089 Pseudocystopteris Ching � 304 Scolopendriurn Adans . � 290 . Pseudodryna ria C.Chr . 160 Scyphofilix Thouars \ gen . dub . ). � 239? Pseudophegopteris Ching 263 Scypholepia J.Srnith � 239 Pseudot ectaria Tard . 339 Scypnopteris Rarin.( gen.dub.) � 239? Psilodochea Presl � 019 Scyphularia Fee 386 Psilogramme Kuhn emend . Underw . 073 Scytopteris Presl � 16) Psilotum Sw . 001 *Selaginella P.Beauv . 008 Psomiocarpa Presl 331 Selaginoioes Seguier (1754) 008 Psygmiurn Presl � 156 Selaginoides Boehm .( 1760 ) 008 Pt eretis Rarin. � 301 Selago P.Br.( 175b) � 008 Pt eridanetium Copel . � 090 Selago Hill( 1757 ) � 006 Pteridella Mett. � 057 Selago Boehm.(1760 � 006 *Pteridiurn Gled. ex Scop.(1760 ) 244 Selago Schur (1866 l � 006 Pt eridiurn Rarin .( 1814) � 096 Selenodesmium ( f'rantl)Co pel . 125 Pt eridoblechnum Henniprnan 408 Selliguea Bory 182 Pteridrys C.Chr. & Chlng 335 Serpyllopsis v.d.Bosch 104 Pteriglyphis Fee· � 307 Sinopteris C.Chr . & Ching 055 Pterilis Rarin. (1819) � 301 Sitobol1urn Desv .( SlJhalm.) � 238 Pt erilis Rafin .(18j0 ) � 096 Sitolobiurn Desv . corr . J.Sn,ith � 23il Pt erinod es O.Kt ze � 301 Sivetes Rarin. � 009 Pt eris L.(1753 ) 096 Solanopteris Copel . 189 Pt eris Scop.(1760 ) � 369 Solenopteris Copel o( sphalrn .) 189 Pt eroneuron Fee � 384 Sorolepidiurn Ctiri st � 356 Pteropsis Desv . � 1 65 Sorornanes Fee � 367 Pt erozoniurn Fee 070 Spathept eris Pr�sl � 035 Pt ilophyllum v.d.Bosch � 140 Sphb erocioniurn Presl 111 Pt ilopteris Hance � 251 Sphaeropteris Bernh .( 18u1 ) 232 Ptychophyllum Presl � 108 Sphaeropt eris Wall .(1830 ) � 354 Pycnodoria Presl � 096 Spha erostephanos J.Smith 281 Pycnolorna C.Chr. 177 Sphaerostichum Presl � 16) Pycnopteris Moore � 369 *Sphenorneris Maxon 256 Py rrosia Mirbel 163 Spheroidea Dulac � 41 0 Pyxidaria Gled . � 117 Spi canta Presl � 399 162 FERN GAZETTE : VOLUME 11 PART 2 & 3 (1975)

Spi cantopsis Nakai � 399 Thyrsopteris Kunze 235 Stachygynandrum P,Beauv . 008 Thysanob otrya· v.A.v.R. 231 Stegania R.Br . � 399 Thysanosoria Gepp 376 Stegnogramma Bl .( 1 828) 279 · Tm eseopteri s Kunze � 002 Stegnogramma Fourn .( 1872) � 068 Tmesipteri s Bernh . 002 Steiropteris (C.Chr .)Pic,Ser . 280 Todea Willd . 029 St enochlaena J .S.m ith 409 Toppingia 0.& I.Deg.& A.R.Srnith � · 263 St enolobus Presl � 38B Toxopteris Trev . 077 St enolepia v.A.v.R. 373 Trachypteris Andre 061 St enotoma Fee (nom .conf'us ,) � 256, Trichi ocarpa (Hook.)J.Smith � 342 257 Trichiogramme Kuhn � 075 Stenosemia Presl 350 Tricqocyclus Dulac � j19 Stibasia Presl � 022 Trichomanes L.( 1753 ) 117 Sti cherus Presl 146 Trich oma(ie s Hill ( 1757 ) = 288 Stigmatopteris )C .Chr . 371 Trichoneuron Ching 327 Stormesia Ki ckx f. � 288 Trichipteris Presl � 229 St romatopteris Mett . 143 Trichopt eris Presl 229 St ruthiopteris Scop . ( 1 754) � 301 Trichosorus Liebm . � 226 Struthi opteris Weis ( 1770) � 399 Trigonospora Holtt . 271 Struthiopteris Willd .( 1bU9 ) � 301 Triphlebia Bak . = 298 Struthopteris Bernh . � 026 Trismeria Fee 064 Stylites E.Amstutz emend . Rauh 01 0 Trisper.mium Hill = 008 Synamtu ia Presl 205 Tristeca P.Beauv . � 001 Synaphleb1um J .Smith � 252 Trochopteris Gardn . = 035 Syngralllma J,Smith 075 Trogostolon Copel . 385 Syngra111matopsis Al ston 071 Synochlamys Fee � 057 Ugena Cav . = 034 Urostachy s Herter = 006

Vaginularia Fee 095 Taeniopsis J,Smith � 093 Vallifilix Thouars = 034 Taeni opteris Hook . � 093 Vandenboschia Copel . = 117 Taenitis Willd . 079 Vi ttaria Smith 093 Tapeinidium (Presl )C.Chr . 258 Tarachia Presl � 288• Weatherbya Copel . � 193 Taschneria Presl � 119 Wibelia Bernh.(1801 ) = 388 Tectaria Cav . 357 Wibelia Fee (1852) = 258 Tectaridium Copel , 343 Woodsia R.Br. 319 Tegularia Reinw . 353 Woodwardia Smith 404 Teleozoma R.Br. � 037 Teratophyllum Mett . 378 Xiphopt eri s Kaulf . 21 1 Thamnopteris Presl � 288 Xyropteris Kramer 259 Thayeria Copel . 161 Thelypteris Schmidel 260 Zaluzanshia Necker 41 0 Thylacopteris Kunze 208 Zaluzanskia Necker � 410 FERN GAZ. 11 (2 & 3) 1975 163

A NOTE ON THE DISTRIBUTION OF ISOETES IN THE CADIZ PROVINCE, SPAIN

BETTY MO LESWORTH ALLEN Fincha la Rana, Los Barrios, Cadiz Province

ABSTRACT Brief notes on the ecology, and distribution in Cadiz Province of lsoetes histrix, I. durieui and I. velata are giveri. /. durieui, although frequent in southern Portugal. has not been hitherto recorded for southern Spain.

When Clive Jermy suggested that I look for lsoetes in my area of the Cadiz province, it stimulated my waning interest in ferns. I had not really bothered about this easily overlooked genus but when I did, an interesting picture unfolded. First, it appeared that /. histrix Bory is an extremely common plant in most parts of this south west corner of Spain, especially in the low hills. Also that it does not seem to be very selective about its habitat, growing in seasonally damp places by streamsides, on mossy rocks in the lowland corkwoods, on grassy slopes often much trampled by cattle; in heavy clay soil in open fields where slight depressions hold water during the winter, and in the firm sand and soft soil behind salt marshes or estuaries (especially near Cadiz). In the hills, too, at 500 m where it becomes a little more selective, this species is common in seepages and in humus-filled crevices between the large rock outcrops which dominate most of the hilltops. Sometimes /. histrix is abundant over a wide area, and on an open hilltop not far from San Roque, at about 240 m, it forms sward-like patches often mixed with Ophiog/ossum lusitanicum, Cicendia filiformis and Crassula tillaea. The soi l is a sandy loam and the surrounding vegetation consists of short windswept Lavandu/a stoechas, Genista triacanthos, Lithodora diffusa. Cistus salvifolius, C. crispus; Satureja salzmannii, Chamaespartium tridentatum, Erodium cicutarium and Ulex parviflorus. All this area is continually damp from about October/November until at least early March, and about May the leaves of /. histrix start disappearing. Practically all the above habitats become very dry and hot during the summer, but occasionally /. histrix grows where the seepage remains damp. Th is is common just above small streams of Rhodendron or Oleander. in light corkwoods. The leaves of these lsoetes are usually much longer (to 28 cm) and more flaccid than normal, but die back for the summer. The megaspores appear to be developed .later than those growing in the wet/dry cycle. This habitat is more where one would expect to find /. durieui. A second species, /. durieui Bory is not nearly so common here as the former and seems to be more frequently found on sloping grassy ground in seepages where there is more shade, again in ground we ll broken by cattle hooves. The distribution seems to be imperfectly known for in the Atlas Florae Europaeae (1. Pteridophyta, p. 31 ), /. durieui is shown for Spain only in the east near the Pyrenees, although it occurs in southwest Portugal not far from the Spanish border. In the Cadiz province (see Alien, 1975) I have found it from 750 m on the summit of a ridge above Algeciras down to about 170 m, but sporadically; also from near the Malaga border at 680 m altitude. lt would appear to require more permanent moisture, shade and possibly higher altitudes on the whole than I. histrix, and most of these areas are damp all through the year, although some have shade for only part of the day. On the summits of the Algeciras hills, mist lies frequently throughout the summer, and it was here that 164 FERN GAZETTE: VOLUME 11 PART 2 & 3 (1975)

I found some plants of I. durieui on a moss-covered rock which was flat and thinly covered with soil. Over this wa ter was seeping and with it were some short grasses, including Poa in ferma, and also Moenchia erecta, Sp ergularia sp., Campanula erinus, Plantago serraria wh ilst Te esdalia nudicaulis grew alongside in a drier place. Other areas are in sm all openings in mi"xed forest of Ouercus suber and 0. faginea mostly_ on· sloping grassy, cattle·trodden ground. Occasionally it grows fairly near I. histrix but so far I have not found them actually growi ng side by side. The most di stant locality within this province, near the Cerro de Jovati, is towards Cartes de la Frontera. These plants were in a fairly open place, again in a seepage but in base-poor soil, and with them we�e the "indicator plants" (for I. histrix ) Cicendia filiformis, Pinguicula lusitanica, Anaga/lis crassula and a few sma ll grasses. Just above the seepage was heathland of short Erica umbe/lata, E. australis, Gal/una vulgaris and Cistus spp. etc. lt suggests that I. durieui may occur more frequently than is generally supposed, especially from 300 or 400 m upwards, in this province anyway. Cattle, deer and goats almost certainly play an important part in the distribution of this species as wel l of I. histrix, the megaspores becoming imbedded in the cloven hooves. Have the hard, black leaf-bases which are present in these two terrestial species some function in protecting the resting stems from damage? A species of Triglochin (T. barrelieri) wh ich grows here in salt marshes as well as inland, and commonly together with I. histrix, has simi lar black and hardened leafbases and it also rests during the summer. The other species which I have seen here is I. velata A.Br. is much rarer and I know of only two places in one locality where it grows. This is an aquatic species, but this locality dries out in the summer; it was growing in a pool on the flat top of a large and isolated rock in full sun. According to the watermark on the side of the pool, the normal water depth would be about 60-80 cm and when full, the leaves of I. velata are submerged and spread more or less in a star-shape. When the water is lower, and this seems" to be the mo re usual state now that the climate appears to be beccrning drier, the leaves lie across the top of the water and are flaccid and bright green. Normally they should be submerged from about November or December to May, exceptionally from September, when the new growth would be in danger of drying out again in December, wh ich is ma inly a dry month. Last year the megaspores were becoming mature in February, probably as there had been little rain in January; in April they were quite mature and greyish. They start to die down as the water dries up, which it does rapidly in May. Mixed with the lsoetes was Catlitriche brutia, and on the edge of the flat side of the pool, in black earth was I. histrix forming a dense fringe with leaves longer than usual. There were a few plants of Op hioglossum lusticanicum on the outer edge, Serapias probably parviflora, Lythrum portula, Allium triquetrum and one plant of Lavandula stoechas. The earth at the base of the pool was not deep and in July and August the heat thrown off into this depression, from the vast white surrounding stone, must be intense. As I write this however, a widespread fire has alre<�dv burnt the su rrounding vegetation and for hundreds of acres surrounding this area, and so it remains to be seen wh at the result will be to thisplan t, and whether the I. histrix and Op hioglossum which were so common amongst the vegetation below, wi ll appear this year after the autumn rains.

REFERENCES ALLEN, B. M. 1975. lsoetes durieui in Spain. Lagascalia (in press). JALAS, J. & SUOMI NEN, J. (eds) 1972. Atlas Florae Europaeae. Distrib ution of vascular plants in Europe 1, Pteridophyta. Helsinki. FERN GAZ. 11(2 & 3) 1975 165

LECANOPTERIS SPJNOSA - A NEW ANT-FERN FROM INDONES IA

A.C. JERMY British Museum (Natural History) London SW7 5BD. and T.G. WALKER Dept. Plant Biology, University of Newcastle upon Tyne NE1 7RU

AB.STRACT Whilst collecting plants in the Latimojong Mnts., Sulawesi (Celebes) the authors fou nd a new species of ant-inhabited fern here described as Lecanopteris spinosa sp. nov. An . account of the morphology and anatomy is given and comparisons are made with other species variously placed in the genera Lecanopteris Reinw. and Myrmecophifa (Christ) Nakai. On grounds of anatomy and rhizome morphology it is argued that the new species is intermediate between these genera thus supporting Copeland's view that they should be united. The highly developed rhizome structure is discussed in relation to the ants that inhabit it.

INTRODUCTION In 1824 Reinwardt described a fern wh ich he called Onychium carnosum, unaware that Kaulfuss (1820) had already used the generic name for another totally different species. No sooner than the name was publ ished Reinwardt became aware of his mistake and in 1825 published a substitute generic name, Lecanopteris. Blume (1828a) elaborated the description emphasizing that the species was distinct from any other Po/ypodium in having a peculiar habit with a swollen rhizome and sari immersed at the reflexed tips of the pinnae segments. Later (1828b) he figured L.carnosa together with a variant which he called L.pumila although the text description of th is new species never appeared. The genus was upheld by some pteridologists (e.g., Presl 1836, Fee 1852 ) whilst · others (e.g. Mettenius 1856; Hooker & Baker 1868, 1874) sti 11 preferred to treat it under a large and variable Polyp odium. Cesati ( 1876) described a related species (now L. deparioides (Ces.) Bak.) under Davallia bel ieving the reflexed leaf tip to be similar to the indusium of that genus. In 1881 however Baker accepted Lecanopteris and described a fourth species (L. incurva ta ) from Sumatra placing it here on account of its soral structure alone, as his specime n (Curtis 24: K) was without a rhizome. Later collections of what is undoubtedly the same species (e.g. Benkoelen, Brooks, s.n. 1920-23; BM) show a typical swollen and naked rhizome. Christ (1897 ), whilst not accepting the generic rank, did however accept the taxonomic concept of Baker (et alia) and he went further and grouped together four other species of Po/ypodium which had similar swollen ant-inhabited rhizomes, under the section Myrmecophila, a name legally instated at generic rank by Nakai ( 1929 ). Whilst the rhizomes of Myrmecophila were ant inhabited and often massive they were not naked but covered with distinct clathrate, dark-centred scales, The sari were superficial, although sometimes deeply immersed lying close to the rachis or costa. This taxonomy wa s maintained although some (e:g. van Alderwelt van Rosenburgh · 1909) went further and included Lecanopteris as a subsection within· section Myrmecophila, when others (e.g. Ching 1940 ) accepted both as genera. Copeland (1905) maintained Lecanopteris (pro gen.) but placed Myrmecophila as a section of Po/ypodium. In 1929 he pointed out the affinity of the former to Microsorium sect. Ph yr1 1atodes and its relationship to Myrmecophila which he proposed (1929: 122) should be united under Lecanopteris, a position he maintained in 1947 and 1960. 166 FERN GAZETTE: VOLUME 11 PART 2 & 3 (1975)

·3

1 cm

IY

1 00 �m

FIGURES 1-4, Lecanopteris �pin osa: 1, diagrammatic sketch of whole plant, (a, growing point; b, advancing front of black pigment zone); 2, portion of rhizome; 3, enlargement of portion of frond showing venation; 4, hairs from rhizome apex. JERMY, WALKER: LECANOPTERIS SPINOSA 167

Holttum (1954) placed the two species of section Myrmecophila fou nd in Malaya in Ph yma todes Presl. [P. sinuosa (Wall. ex Hook. ) J. Sm.; P. crustacea (Copel.) Holtt.] on the grounds that they "have all the characters ... [of Ph yma todes] apart from the swollen rhizome". They do not however have the characteristic knobbly, branched paraphyses amongst the sporangia as the other species of this genus, e.g. P. scolopendria (Burm. ) Ching. Lecanopteris, with its distinct sori and clathrate scales, Holttum keeps as a separate genus. · The finding of an ant-fern in Sulawesi with a swollen, naked rhizome but with fronds with the texture and sori of L. sinuosa places it intermediate between sections Lecanopteris and Myrmecophila. Although in the two genera the position of the sori

TAB LE 1: Species of Lecanop teris discu ssed in this paper.

Rhizome naked; soru s only slightly Rhizome with clathrate scales; sorus immersedat pinnae or segment ends immersed close to costa of ± simple which are ± reflexed (= section fronds or closeto costu le of pinnae Lecanopteris) (= sectionMyrmecophila)

L. carnosa (Reinw. ) Bl. L. lomarioides (Brause) Baker L. davallioides v.A.v.R. (incl. L. crustaceaCopell L. deparioides (Ces.l Bak. L. mirab ilis (C. Chr.) Ching (incl. L. curtisii Bak.) L. sarcopus de Vr. & van Teyss. L. incurvata Bak. L. sinuosa (Wall.) Copel. L. nieuwenhuisii Christ* (incl. L. pumila Bl. ex Copel. ) L. philippensis v.A.v.R.* L spinosa Jermy & T. G. Walker

* rhizome not seen varie� and this is correlated with the dissection of the frond, the vascular supply to it and its general structure is similar. The species examined by us are listed under the two sections in Ta ble 1. In all, the maturing fertile sporangia are immersed in a mat of long setae topped by an abortive sporangium which often breaks away leaving a curious cup-shaped head and which, although possibly homologous with the branched paraphyses of other Microsorioid genera, should not be confused with them. The rhizome architecture varies from the more massive irregular growth pattern in L. crustacea through L. spinosa, both without scales, to the regular L. sinuosa with scales; L. mlrabilis is irregular and expansive with often an almost naked upper surface; according to Copeland (1929) L. sinuosa has a Papuan form (not seen by us\ with a glabrescent rhizome. Except for L. da vallioides and L. nieu wenhuisii, where we have seen no rhizome material, all the species in section Lecanopteris have uniseriate, occasionally branched, trichomes on the epidermis of the young rhizome. In anatomical structure both the rhizome and frond of L. sp inosa are similar to L. carnosa and L. sinuosa (cf. Yapp 1902). lt is seen that L. sp inosa is intermediate between the two sections as hitherto defined and is further proof in support of Copeland to unite Myrmecophila and Lecanopteris under the latter. If the degree of organisation and adaption of the rhizome is significant, L. sp inosa would be the most advanced member of section Lecanop teris.

DIAGNOSIS AND DESCRIPTION lecanopteris spinosa Jermy & Walker sp. nov. (fig 1 et sequ.) Planta myrmecophila epiphytica. Rhizorna glabrum ventricosum 2.5-4 cm crassum, ram is brevibus 161l FERN GAZETTE : VOLUME 11 PART 2 & 3 (1975)

FIGURES 5-6, Lecanopteris spinosa: 5, plant in situ x )(.; 6, sectioned rhizome showing galleries (arrow intersection indicates ants removing pupae). JERMY, WA LKER: LECANOPTE RIS SP INOSA 169 lateralibus, porticus ad apicem capiens, initio subv iride pruinosum, demum nigrum coriaceum, spin is ad 6 mm conicis attenuatis. Frondes exstipitatae, bifariae, usque ad 30 x 5 cm (frondes steriles saepe 6-8 cm longae); rhachis flavovirens; la mina frondum sterilium simplex, frondum fertilium pinnatifida lobata, apices loborum rotu ndatis; venae anastomosantes, areolae venulis liberis clavatis includentibus. Sori immersi uniseriati prope rhach idem, sporangia initio sporangibus sterilibus elongatis obtecta at sine paraphysibus ut in speciebus nonnullis Microsorii. Sp orae monoletes (56-) 69-(82.5) Jlmlo ngae luteolae laeves. . Typus Indonesia, Sulawesi, Latimojong Mnts, SW slopes of Mnt. Rantemario, below Buta Jakke above base camp; on Platea la tifolia Bl. in mixed oak-podocarp forest, at 1950 m alt., 10 November 1969, A. C. Jermy 7609; holotypus in BM; isotypi in BO, GH, L. Paratypus: Loco citato, T. G Wa lker T12179 (in herb Walker). Plants epiphytic, found on the upper branches (3-6 cm thick) of trees 15-20 m high. Primary rhizome 2.5-4 cm thick, the advancing front somewhat flattened and adpressed to the substrate and budding off lobes which develop into lateral branches 2.5-5 cm long, 1-3 cm th!ck, and which on maturity swell to press against each.other (see fig 2). In all the plants studied the main axis of the rhizome soon orientated itself across the branch and gradually encircled it. When the first complete tu rn has been achieved the m a in rh izome continues to grow in the same direction on top of the old dead, leathery rh izome. Thus a "ball" of rhizome would result often three or four stems thick, each growing epiphytically on itself and in which there would be usually on ly one advancing apex although occasionally a side branch would continue growth for a time. lt is possible that such "balls", found up to 25 cm across, could contain more than one plant but this seems unlikely. Living rhizomes are pale green with a pronounced glaucous bloom on the younger parts which dries as flat, plate-like scales of wh ite wax. Scales absent but the very young epidermis is sparsely covered with simple, very rarely branched, caducous trichomes of four to eight cells uniseriately arranged, the terminal cell'of which is not obviously glandular. Epidermis minutely papillate with scattered multicellular spi nes, 4-6 mm long, narrowly conical, or cylindrical and tapered with a swollen conical base, occasionally consisting of th is only, green and flexuous often turning black before the rest of the epidermal tissue. Green rhizome becoming black (see para. on anatomy) as fronds absciss; the tissue eventually dying as the outer layers become horny or leathery. The meristematic tissues of the lateral branches die at this stage and the invaginated apex is often perforated as the tissues shrink (see below). Fronds, in two ranks on the main rhizome, not normally seen on the lateral branches although some are initiated there; for the most part initiated on the me ristematic front of the rh izome which develops, as the leaf ages, into a conical phyllopodium, stipe virtually absent, articulate to rhizome, the fronds falling as the rh izome blackens. Lamina up to 30 x 5 cm (often only 6-8 cm long in sterile fronds ), linear-oblanceolate tapering below to form in the lower 2-5 ems a wing along the rachis, apex rounded, often imperfect ang emarginate, margi n entire in small sterile fronds to undulate or ± regularly lobed j- to the rachis in fertile fronds (see fig 1 ); texture coriaceous, dull, pale green, midrib distinct, yellowish green, raised on the adaxial side and with a narrow ridge on the abaxial surface, the veins conspicuous on drying, anastomosing to form areolae with included free veinlets (see fig 3). Sori in one row close to and either side of the rachis usually in the upper � of the leaf only; circular on flat immersed receptacles which cause projections on the upper surface; sporangia protected in the early stages by long-stalked abortive sporangia which overtop them and whose heads touch one another to form a compact layer (see fig 25). Spores monolete, pale yellow (56-) 69 (-82.5) Jlm long (in a sample of 50), sporoderm smooth under the SEM with only the occasional papilla in the region of the suture. 170 FERN GAZETTE: VOLUME 11 PART 2 & 3 (1975)

14

11 12

FIGURES 7- 15, Lecanop teris spinOS/1 anatomy: 7, epidermis of rhizome (x 65); 8, lower epidermis of frond (x 65); 9, upper ep iderm is of frond (x 65); 10, LS. young dehiscence region of ·frond base (x 45); 11, LS old frond base (x 45) ; 12, LS old rh izome at junction of green and black areas (x 65); 13, LS young rhizome, boundary between lower water storage tissue and upper parenchyma arrowed ( x 55); 14, delim iting region arrowed in fig 13 at later stage just prior to breakdown of water storage tissue of rh izome (x ·70); 15, TS rhizome showing galleries (cross hatchlild) and vascular system. JERMY, WALKER: LECANOPTERIS SPINOSA 171

ANATOMY 1. The root The roots are short, black and wi ry and cling to the substrate and when developed in the chambers produce copious root-hairs. The xy lem is diarch and there is a particularly well defined pericycle (fig 26). Abutting onto the endodermis is a ring of cells with heavily thickened wa lls about seven cells in depth, except opposite the protoxylem points where it is interrupted. Th is is identical to the root structure of L. carnosa as described by Yapp (1902).

2. Therhizome In the young parts the spines, which are such a conspicuous and permanent feature of th is plant are green and flexible when young, consisting of cells with living contents and having lignified wa lls. As they age the cells lose their contents (fig 24) and the spines become rigid. They first blacken at the tip and then ultimately throughout, this blackening frequently preceding that of the rh izome discussed later. The rhizome epidermis consists of thick walled cells and is devoid of stomata (fig 7). OJerlying the cuticle is a very thick transparent, wax, becoming white on drying seen under the SEM as irregular flakes which harden to the surface when the rh izome blackens. As in some other species of Lecanopteris e.g. L. carnosa, the rhizome blackens with age and a curious feature in L sp inosa is that frequently the boundary betwe en green and black areas is absolutely sharp as if drawn by a pen. Although, as ageing occurs, there is a gradual browning of the underlying tissues, the black appearance is entirely caused by the deposition of dense melanic substances in the epidermal layer only (fig 12). Immediately beneath the epidermis there is a layer a few cells deep which is lignified and internal to this is parenchyma. In the centre of the young parts of the rh izome the parenchyma fo rms a very distinctive water storage tissue which can be distinguished by the naked eye by its lighter colour. In this tissue the cel ls are much larger than those of the surrounding parenchyma, are elongated vertically and have a glistening translucent appearance due to their high water content. lt is the breakdown of this tissue which leads to the formation of the cavities. The water storage tissue is of short duration, rarely being found intact more than about two centimetres behind the apex of the main stem of the branches. This agrees with the situation as described for L. carnosa by Yapp but although he described some of the anatomical details their possible importance in the formation of the cavities appears to have eluded him as he believed that only very young plants (wi1 ich he did not possess) would provide the answer. The innermost one or two layers of the parenchyma bordering onto the water storage tissue are not especially prominent at first, being thin wal led and tending to be aligned with their long axes in the same plane as the ce lls of the water storage tissue (arrowed in fig 13). Later on, however, these parenchyma cells divide to form a very conspicuous delimiting layer from which finger-like processes project between the outermost cells of the water storage tissue (fig 14). Th is later stage is illustrated by Yapp (1902). The cells of this delimiting layer become rapidly and progressively more suberized and when this process has reached a certain stage the water storage tissue also develops a very small amount of suberization. At th is point the del imiting layer apparently cuts off the once more or less intimate contact between the water storage tissue and the ground parenchyma of the rh izome, with the subsequent disintegration of the water storage tissue and the resultant space forming a chamber lined by a layer 172 FERN GAZETTE: VOLUME 11 PART 2 & 3 (1975)

�� 23

25 26

FIGURES 16-26, Lecanopteris sp inosa anatomy: 16-23, TS rachis from base towards apex, xylem as solid black, endodermis as broken line (x 5.75); 24, TS mature spine of rhizome (x 55); 25, LS young sorus (x 90); 26, TS stele and inner cortex of root (x 400). JERMY, WALKER: LECANOPTERIS SPINOSA 173 of dark suberized cells. If a young apex is sectioned the development of the delimiting layer, its progressive suberization and the breakdown of the water storage tissue can be seen as a continuous process occurring over a distance of 0.5-1 cm. Hence the chambers or galleries are constantly being extended, keeping pace with the elongation of the rhizome. The apices of young rhizome lobes or of the lateral branches are frequently invaginated and have the appearance of a "pore". On dying the parenchymatous cells at the base of this pore shrink and break down thus linking the chamber with the exterior. lt is through such natural openings that ants gain entrance to the galleries. The cells lining these openings are simi lar to those in the galleries suggesting that the opening has been formed in much the same way as the galleries themselves. In actively growi ng portions of the rhizome there is a "lip" on the leading edge which consists of young dividing cells forming the meristem. lt is in th is region that the leaf primordia and the spines arise (marked a in fig 1 ). This extends along the growing margin and no single apical cell was detected, although it would be exceedingly difficult to find if it did exist without microtoming a considerable number of sections. Toward the sides as the rhizome matures, or in the lateral branches of limited growth, the edges tend to be more or less rounded, the parenchyma comprising the bu lk of the rhizome being fully expanded and enclosed in a layer of lignified cells on both upper and lower surfaces and bounded by the epidermis. Further back in the blackened areas, all the cells become orange-yellow and spongy in texture. The rhizome is permeated with a network of vascular strands which are embedded in the parenchyma above and below the galleries (see fig 15) but which are absent from the water storage tissue where th is is present. The individual strands are very similar in structure to the vascular supply of the root (fig 26) except that immediately outside the endodermis there is at most a single layer of thickened cells. The pattern of the galleries tends to be rather more complex in th is species than Lecanopteris. in most others of Here they tend to be on three levels wit.h a major gallery running along the entire length of the rhizome and giving off frequent chambers above and below which interconnect with one another. The layout is not so clear as in some other species such as L carnosa and L. deparioides where the rhizome can be likened to a series of flasks joined together, each bearing a frond and having one or more chambers within. In L. sp inosa the fronds tend to be somewhat sparse and the rhizome not organised into units joined together, other than the lateral branches. According to the position of the cut, at any part of an opened rhizome there may be seen three lines of chambers or else large irregular spaces where all the chambers meet (see fig 1 ). Thus there is a continuous network of ga lleries and chambers throughout the plant as in all of the Asiatic ant-ferns. Th is is in contrast to the New World species of So/anop teris, described in detail recently by Hagemann (1969) and Wagner (1972 ), in which the rhizome is dimorphic and .the inflated areas are produced on short shoots.

3. The frond The upper surface of the lamina is devoid of stomata unlike the lower epidermis which is richly provided with them (figs 8-9) the mean length of the guard cells being 46pm (range 41-52 pm in a sample of 50). The photosynthetic region consists of several layers of cells which are densely filled with chloroplasts and form a spongy mesophyll. Yapp ( 1902) reports the presence in L. carnosa of a single layer of large colourless cells immediately beneath the upper epidermis which possibly forms a hypodermal water reservoir but in L. spin osa the situation is somewhat variable, even in the same section. In some areas the layer is clearly present, containing conspicuous cells which 174 FERN GA ZE ITE:VO LUME 11 PA RT 2 & 3 ( 1975)

completely lack chloroplasts, wh ilst in other areas these cells are less obvious and contain some chloroplasts, although fewe r in number than in the spongy mesophyll. The fronds are initiated at the edge of the rhizome and are completely naked throughout life. Each is borne on a prominent cone-shaped leaf base (phyllopodium) and there is a well-defined dehiscence region present from the beginning which results in the old fronds being shed cleanly. This region consists of a layer of rather irregularly shaped, but more or less isodiametric cells some 12 rows deep, which contrastwith the more elongated cells on either side, i.e. in the base of the stipe and in the top of the phyllopodium. The cell walls of the dehiscence layer, with the exception of the uppermost 2 or 3 rows, start to thicken early in development and can be seen even when the young frond has just unfurled (fig 10). Lignification proceeds and ultimately a layer some 20 or more cells thick caps an old leaf base from which the frond has been shed (fig 11 ). Dehiscence occurs by tearing along the uppermost two or three rows of unlignified cells, and their torn remains may be seen in section. This behaviour contrasts with that reported by Phillips & White (1967) for representatives of several genera belonging to the Polypodiaceae in which a specialised dehiscence layer was by no means so conspicuous. Six or seven vascular strands are present in the leaf base and traverse the dehiscence layer into the bottom of the rachis, where they become orientated into two vertical rows (fig 16). Theup per two strands nearest the adaxial side of the rachis tend to be more conspicuous than the others and arc elongated transversely. They approach one another more closely at successive levels up the rachis and quickly fuse to form a very prominent trace which then persists through the remaining length of the rachis. This behaviour is in contrast to that of the other strands which gradually decrease in number by a process of fusion and finally disappear near the tip, leaving only the conspicuous trace noted above (·figs 16-23). The sori are produced in a single row on either side of the rachis and are deeply impressed (fig 1). 1l1ebase of the sorus is richly supplied by vascular tissue which is delimited from the other tissues of the lamina by a single continuous layer of very thick walled cells on the under side only and abutting on to the endodermis. One might speculate as to whether or not this is a device whereby the supplies carried in the vascular strands are diverted to supplying the sorus rather than the underlying tissues of the lami na. There is no mention of this structure in Yapp's account of L. carnosa and we fa iled to find it in ou r sections of L sinuosa and L. deparioides.

ANT-FERN RELATIONSHIP The rhizome galleries of L. spinosa were inhabited by a species of Crematogaster. Specimens have been deposited in the Department of Entomology, British Museum (Natu ral History ). lt is possible that other genera or species of ants were commensal in the rh izome but we have no evidence of this; Gomez (1974) found at least three genera coexisting in So/anopteris brunei (Werkle ex Christ) Wagner. Ants ga in access to the chambers through the apical cavities of the lateral branches and we have no ' evidence that the ants eat their way into the chambers as do Azteca sp. into So/anop teris (Gomez, I.e. ) although they may physically remove dead cells and other detritus that accumul ates in the aperture. A full discussion on the evolution of the interaction of lridomyrmex myrmecodiae with l.ecanopteris sinuosa and other mymecophytes in Sarawak is given by Janzen (1974). The followi ng observations are give'n in support of his ideas. JERMY, WALKER: LECAN OPTERIS SP IN OOA 175

The affect of the plant on the ant Apart from shelter ants may absorb moisture from the living tissue either directly or indirectly. The outer epidermis of L. sp inosa is hard and impervious to water but rain water could perculate through open phyllopodia or apical "pores" and become absorbed on the spongy dead parenchyma thus maintaining a moist atmosphere within the chambers. To "what extent the Lecanopteris attracts other arthropods which in tu rn are captured by Crematogaster, by being a source of food or shelter was not observed but little evidence of such is seen on dried material now before us. On analysis, the heaps of debris found rare�y at the ends of the galleries consisted mainly of heads (and leg and mouth-parts) of Crematogaster and no other specifically recognisable remains except coleopteran (? ) larva/castes and elytra and many fragments of insect wing and limbs. . . That the fern offers food in the way of highly nutritious spores or young or abortive sporangia has been suggested by several workers (Holttum _1954a, b,; Janzen 1974). Certainly in many cases the whole sorus is removed, presumably by some grazing animal and often the lamina beneath the sorus is also eaten away. As Holttum (1954b)' reports for L. sinuosa, the paraphysis-like abortive sporangia are full of globules that have the appearance of oil bodies, although we found these difficult to stain with Sudan blue/red. In all the species of Lecanopteris studied by us most of the abortive sporangia! heads had been lost at the time of sporangia! dehiscence leaving a characteristic broken cell at the top of the seta. Thismay be due to natural fall off or to grazing by ants but it must be admitted that this type of "paraphysis" is seen in many Microsorium species not associated with ants. We have not investigated the chemistry of the rhizome tissue and it is possible that the large parenchymatous cells described above contain sugars similar to those in Solanopteris (Gomez, I.e. ). Ants are certainly found in the green succulent rhizome.

Theeffect of the ant on the plant Holttum speculates that in return for shelter provided by the plant the ants bring in mineral substances from the ground and also some nitrogen in their excreta. No evidence of the former was fou nd in L. sp inosa and no mineral carton was seen associated with the species as is often the case with other myrmecophytes, e.g. Hydnophytum and Myrmecodia (Janzen 1974). Fern roots were seen to penetrate the cavities, usually through broken down phyllopodia, and once inside developed copious root hairs; whether they absorb anything other than water has yet to be proved. Another hypothesis is that frequently put forward, and with good evidence, in connection with some flowering plants which also show an association with ants, namely that the ants protect the plant from attack by insects. Our limited observations suggest that first, the ants tended to flee rather than attack when the plant was touched and second, fronds and sori ·showed signs of being eaten presumably by a herbivorous insect other thanthe ants themselves. In L. sp inosa and in the genus as a whole the young fleshy parts of the rhizome are devoid of scales or very poorly furnished with them wh ilst at the same time occupyingan exposed position in the tree canopy. Certain parallels exist between this situation and that discussed in some detail by Janzen (1966) for the "swollen thorn" acacias of Central America. lt is certain that in cultivation, slugs wi ll seek out Lecanopteris rhizomes in preference to other fern species. Th is was demonstrated at Newcastle when 39 sporelings of L. mirabilis were raised, in the absence of ants, dispersed among several propagating frames and intermixed at random among several hundred sporelings belonging to a large number of genera. In every case the Lecanopteris sporelings were irrevocably damaged by slugs 176 FERN GAZETTE: V OL UME 11 PART 2 & 3 (1975)

whilst those of the surrounding species had hardly been affected. lt is unlikely however that slugs would be predators in nature but the observation establishes the palatability of the rhizome tissue wh ich in nature is untouched. Although the rhizome is green, little or no C0 exchange can take place through 2 the waxy cuticle. The presence of actively ·respiring ants in the chambers of the green rhizome could then be an advantage.

ACKNOWLEDGEMENTS This material was collected on a British Museum - Newcastle University - Kew expedition carried out with the co-operation of the National Biological Institute, Indonesia. ACJ would like to thank the Trustees of the Museum and TGW the Nuffield Foundation for helping to finance this expedition as part of larger research projects. We would particularly like to thank Dr Mien Rifai, Director, Herbarium Bogoriense for his help as liaison officer both in Sulawesi and elsewhere in Indonesia.

REFERENCES VAN ALDERWERELT VAN ROSENBURGH, C.R.W.K. 1909. Handbk. Malay. Ferns. Batavia. BAKER, J.G. 1881. On a collection of ferns made by Mr Curtis in the Malay Islands and Madagascar. J. Bot., Land. N.S. 10: 366-368. BLUME, C.L. 1828a Enum. PI. Jav. 2: 120. B LUME, C. L. 1828b. Fl. Javanica. t. 94, B ruxelles. CESATI, V. 1876. Felci e specie nei gruppi affini raccolte a Borneo dal S.O. Beccari Atti Accad. Se. fis. mat. Napoli. 7(8): 1-42. 1940. 5: CHING, R.C. On natural classification of the family "Poly. podiaceae". Sunyatsenia 201-306. CHR 1ST, H. 1897. Die Farukrauter der Erde. Jena. COPE LA ND, E.B. 1905. Th e Polypodiaceae of the Philipp ine Islands. (Bureau Govt. Labs. Pub!. 28:) 1-139. COPE LA ND, E.B. 1929. The oriental genera of Polypodiaceae. Univ. Calif. Pub/. Bot. 16: 45-128. COPELAND, E.B. 1947. Genera Filicum. Waltham, Mass. COPE LAND, E.B. 1960. Fern flora of the Philippines (M onogr. Nat. Inst. Sci. Tech. No. 6) 3: 498-500. Manila. FEE, A.L. 1852. Mem. 5, Genera Filicum, 259. Paris. GOMEZ, P. L.D. 1974. The biology of the potato-fern Solanopteris brunei. Brenesia 4: 37-59. HAGEMANN, W. 1969. Zur Morphologie der Knolle von Polypodium bifrons Hook. und P. brunei Werckle. Mem. Soc. Bot. France 1969: 17-27. HOLTTUM, R.E. 1954e. Flora of Malaya 2, Ferns: 188-191, 208-210. Singapore. HOLTTU M, R.E. 1954b. Plant Life in Malaya. London. HOOKER, W.J. & BAKER, J.G. 1868, 1874. Syn. Filicum, eels 1 & 2. London. JANZEN, D.H. 1966. Coevolution of mu tualism between ants and acacias in Central America. Evolution 20: 249-279. JANZEN, D.H. 1974. Epiphytic myrmecophytes in Sarawak: mutualism through the feed ing of plants by ants. Biotropica 6: 237-259. �AULFUSS, G.F. 1820. Jahr. Pharm, Berl. 1820: 45 . . METTENIUS, G. 1856. Farngattungen, I Polypodiacea, 102, Frankfurt am Main. NAKAI, T. 1929. Notes of Japanese ferns VIII. Bot. Mag. , To kyo 43:6.

'i>HI· LUPS, D.A. & WHITE, R.A. 1967. Frond articulation in species of Polypodiaceae and Davalliaceae. Amer. Fern J. 51: 78-88. PRESL; C.B. 1836. Te ntamen Pteridographiae seu genera filicacearum. Prague. REINWARDT, C.G.C. 1824. Syllog. pi. nov. 2: 3. REINWARDT, C.G.C. 1825. Flora, Jena 3 Beibl., 48. WAG NER, W.H. 1972. Solanop teris brunei, a littl&-known fern epiphyte with dimorphic stems. Amer. Fern J. 62 : 33-43. YAPP, R.H. 1902. Two Malayan "myrmecophilous" ferns, Po/ypodium (L ecanopteris) carnosum (Biume), and Polypodium sinuosum Wall. Ann. Bot. 16: 185-231. FERN GAZ. 11(2 & 3) 1975 177

DRYOPTERIS TYRRHENA NOM. NOV. - A MISUNDERSTOOD WESTERN MEDITERRANEAN SPECIES

C.R. FRASER-JENKI NS Radley College, Abingdon, Berks., England T. REICHSTEIN I nstitut f�r Organische Chemie der Universitat, 19 St Johanns-R ing, CH-4056 Basel, Switzerland G. VIDA Dept. of Genetics, Eotv8s Lorand University, Muzeum k8rut 4a, H-1088 Budapest, Hungary

ABSTRACT A rare west mediterranean fern first found on the Sierra Nevada (southern Spain), and described as Aspidium nevadense Boissier (1938) has since been reduced by most botanists to a subsp. or var. of different representatives of either the Dryopteris filix ·mas or D. villarii complexes. In our opinion it should be recognised as a good species, and as· D. nevadensis is unfortunately no longer available we propose the name D. tyrrhena. In its morphology it is intermediate between members of the D. filix·mas and D. villarii complexes and we assume that it may once have arisen by chromosome doubling from a diploid hybrid of D. abbreviata x D. vil/arii subsp. pallida. Chemical results agree well with this hypothesis but final cytological proof must await further ex perimental work. We also describe two important back crosses: D. x sardoa, a natural triploid hybrid of D. abbreviata x tyrrhena and an experimental triploid of D. tyrrhena x D. villarii subsp. pallida. The latter showing c. 40 bivalents in meiosis in agreement with the above hypothesis. We have collected or seen specimens of D. tyrrhena from the Sierra Nevada, the Tyrrhenian Islands (Corsica, Sardinia, Elba, Capraia) and from Italy (Riomaggiore, Liguria) from altitudes near sea level to c. 2500 m. Its disjunct distribution in isolated places, often in caves or deep crevices in silicate rock, suggests that it is an old rei ict species.

INTRODUCTION Many species of the genus Dryopteris are critical taxa, often difficult to differe'ntiate. In Europe (excluding the Azores) three complexes are known, and only after the introduction of cytology to fern taxonomy by Manton (1950), was a better understanding of their classification possible. Today the following species are accepted as members of these complexes. A. The D. carthusiana-dilatata complex (formerly known as D. sp inulosa complex) comprising: D. aemula (Ait.) 0. Kuntze, D. assimilis S. Walker, D. carthusiana (Vill.) H.P. Fuchs, D. cristata (L.) A. Gray and D. dila ta ta (Hoffm.) A. Gray.* This is not involved in this study. B. The D. filix-mas complex comprising 1. D. abbreviata (DC.) Newm., * diploid; 2. D. caucasica (A.Br.) Fraser-Jenkins & Corley (1972). diploid, known so far only from the Caucasus, Tu rkey and Persia; 3. D. filix-mas ( L) Schott, allotetraploid and 4. D. borreri Newm., * apogamous with diploid and triploid cytotypes. Manton (1950) has shown that D. filix-mas contains two genomes of D. abbreviata and according to

*In this article we use Heywood's nomenclature of Flora Europaea (1964) despite the fact that some names are illegitimate. This is true for D. abbreviata (pC.) Newm. (see Fraser-Jenkins & Jermy, in press ), D. borreri Newm. (see Holub 1967), and D. dila tata (Hoffm.) A. Gray (see Jermy 1969 ). 178 FERN GAZETTE: VOLUME 11 PART 2 & 3 (1975)

Fraser-Jenkins & Corley (1972) the two other ones most probably come from D. caucasica. As these two diploid ancestors are morphologically somewhat similar it is understandable that all these three species can sometimes be very difficult to tell apart. C. The D. villarii complex, growing nearly exclusively on limestone. This is today treated as one species, following Heywood (1964) and divided as follows: 1. D. villarii (Bell.) Woynar subsp. villarii, the alpine diploid; 2. D. villarii subsp. pallida (Bory) Heywood, a diploid, morphologically rather variable taxon from low altitudes around th e Mediterranean Sea which can further be separated into different varieties of which we mention for the present study only var. ba/earica, originally described as D. rigida (Sw. ) A. Gray subsp. australis (Ten.) C. Chr. var. balearica by R. Litardi'ere (191 1: 23, see also Knoche 1921 : 252). 3. "The tetraploid taxon" (Widln et al. 1971 ). This is the only form of D. villarii found so far in Great Britain (Manton 1950; Gilbert 1966). lt is also distributed on the European Continent from Spain to the Balkans (Vida 1969) mainly above 800 m. lt is probably an allotetraploid derived from the two mentioned diploids (Manton 1950, 1961; Panigraphi 1965) but final proof for the degree of affinity between the three subspecies is still missing. The morphology of the tetraploid is more or less intermediate between the two diploids but nearer to subsp. pallida. Its unequivocal differentiation from the two diploids is difficult and without cytological analysis, sometimes impossible.

METHODS For meiotic counts pinnae with unripe sporangia were fixed in freshly mixed absolute alcohol: glacial acetic acid (3:1) either in the field or from cultivated plants. After 24 hours the liquid was replaced by new mixture kept (wherever possible) at 0°C and the fixings sent to Budapest by air-mail. If th is was not possible the fixings were transferred to 70% aqueous alcohol and kept at room temperature. The samples were kept at Budapest at -15°C until examination. Staining, squashing and the making of permanent preparations was done following Manton (1950: 293-299). Root tips for meiotic counts were treated in 0. 1% aqueous colchicine for 8 hours at 4° C (if th is was not possible 2 hours at 20°C), dried quickly on blotting paper and fixed as mentioned above. Before staining the root tips were softened with snail enzyme (Faberge 1945, Ray & Manton 1965) or commercial "cellulase 50,000" (a preparation from Aspergillus spec. by courtesy of "Schweizerische Ferment AG.", Base!), c. 2% in 1% aqueous acetic acid. Root tips which had been stored in 70% alcohol were first transferred for 1 hour to 1% aqueous acetic acid before treatment with any of the enzymes to eliminate the alcohol which otherwise inhibits the action.

THE TAXONOMIC POSITION AND RANK OF ASPIDIUM NEVADENSE BOISS. We now report on a European fern whose morphology is intermediate between the members of the D. filix-mas and the D. villarii groups. lt i.s therefore understandable that its proper classification caused difficulties even to expert pteridol ogists for more than 130 years. We refer to Aspidium nevadense Boissier (1838) found by that author in the Sierra Nevada (Southern Spain at c 8000' (=c 2600 m) 1 ). We were able to see the type (in G, see fig 2-3) and to also collect consepecific living material (plants and spores) from different local ities (see below and fig 1), now in- cultivation for detailed studies including cytological control. From th is it is evident that Boissier was right in

1 Prov ided Boissier used French feet (pied de Roi 1' � 32.4 cm: we thank Mr F. Badr{, Paris, for information) as did Linnaeus (see Stearn 1966 : 112- 113). FRAS ER-JENKINS, REICHSTEIN, VIDA: DRYOPTE RIS TY RRHENA NOM. NOV. 179

FIGURE 1. Dryopteris tyrrhena in situ, Corsica, Calanche de Piana: The colony from which TR-3562 was collected. In the left lower corner are leaves of Cyclamen neapolitanum. (Ph oto. H. & K. Rasbach 4. 5. 197 1 ).

describing the plant as a distinct species but, due to its intermediate morphology, later botanists reduced it to a subspecies, variety or forma of either D. villarii (e.g. Mettenius 1856-8, Moore 1858, Milde 1867, 1868 ) or of D. filix -mas (see synonymy ). lt was also a reason why Litardihe (1924: 122-126) suggested the treatment of D. villarii as subsp. of D. fil ix-mas, a suggestion obviously not accepted by later authors.

Nomenclature According to current nomenclature Aspidium nevadense Boiss. is a member of the genus Dryop teris Adanson. But according to the Code (Stafleu et al. 1972) it is unfortunately not possible to use the combination D. nevadensis (Boiss.) as Eaton (1878) described a different (American) spec ies as Aspidium nevadense. This was of 180 FERN GAZETTE: VOLUME 11 PART 2 & 3 (1975)

course an illegitimate name but Baker (1891 ) published for this plant the legitimate new name Nephrodium nevadense which was transferred to Dryop teris by Underwood (1893 ). The valid name for the American species under Dryop teris is therefore D. nevadensis (Baker) Underw. lt is true that Christensen ( 1905: 281) suggested the nam·e D. oregana C. Chr. for it, but this must be considered illegitimate for the above reasons. Despite the fact that the plant in the meantime has been transferred to another genus as Th elyp teris nevadensis (Baker) Clute (see Morton 1958), we had to choose a new name for Aspidium nevadense Boiss.

TAXONOMIC TREATMENT ·Synonymy and typification2 ·Dryopteris tyrrhena Fraser-Jenkins et Reichstein, nom. nov. Aspidium nevadense Boiss. ( 1838: 93-94) non Dryop teris nevadensis (Baker) Underw. (1893: 113). _ Aspidium rig/dum "forma pinnatisecta" Milde (1867: 127), nomen invalidum. _ A. rig/dum var. pinnatlsectum Milde (1868: 362).

_ A. rigidum var. nevadense (Boiss.) Luersen (1889: 408,410 fig 150a, a pinnal.

_ A. rigidum "Rasse" nsvadense (Boiss.) Ascherson in Ascherson & Graebner ( 1896: 30, 1913: 44). _ Dryopteris rig/de var. nevadense (Boiss.) C. Christensen (1906:84). _ Nephrodium rigidum subsp.nevadense (Boiss.l Rouy (1913: 409,footno te). _ Dryoptsris filix-mas subsp.rigida var. nevadsnsis (Boiss.) Litardiere (1924: 123). _ D. villarii var. nsvadsnsis (Boiss.) Fiorl (1943: 113 gives a good description but wrong figure with petiolated pinnules, correspondingto subsp.palli da l. Aspidium fifix-mas var. glandulosum suet. non Milde (1867: 123); see below. Dryoptsris filix-mas var. glandulosa suet. pro parte, non (M ilde) 8 riquet ( 1910: 10-11). Nephrodium fil ix-mas "race" rigidoformis auct. parte, non Rouy (1913: 408). Dryopteris litardisrei sensu Rothmaler (1945: 94-95, pro hybr. D. palsacea x pallida l partim, excluso typo. Type: Aspidium nevadense Boiss. (1838) Sierra Nevada, southern Spain, c. 8000' leg. Boissier April, 1837 (G), see figs 2 and 3.

Description of D. tyrrhena Both Boissier's (1838) diagnosis and Milde's (1868: 362) description are excelle nt. Nevertheless the plant has frequently been confused since and we therefore give a detailed description, listing some characters in Table 1 which can be used to distinguish D. abbreviata, D. filix-mas and D. villarii subsp. pallida from D. tyrrhena; these also show the intermediate position of D. tyrrhena in some of these characters. Diagnosis: A D. filix-mas differt: 1. lamina ad basim versus.sa epissime minus angustata; 2. segmentis secundariis (minoribus ad fines frond is pinnarumque exemptis) ± walibus, saepe magis distantibus, basi saepe paulum angustatis mi nusque confluentibus, apice ± rotundatis 0.5-1 mm longe acute denticulatis; 3. lamina indusiis rachidique inclusis utraque pagina pills glanduliferis brevibusdense vestita. Perennla1 1Jiant, .± evergreen if not completely damaged by frost. Rhizome ascendent up to c. 2 cm thick, covered with scales at the apex, slowly branching, thus sometimes forming colonies up to 1 m diameter in suitable' places. Ripe fronds similar to D. filix·mas but only c. 10-50 (-60) cm long and 4- 10 (-16) cm wide. Petiole c. 1/4,-3/4 as long as the lamina. Petiole and rachis like the rhizome densely covered with pale strawcoloured or light reddish brown, finely striped scales. Scales at the base of the petiole up to c. 1-2 cm long and 4-6 mm wide (fig 91. Lamina 1-2 pinnate, deltoid·lanceolate in outline (figs 2 & 4 ). Lowest pair of pinnae r�trely the longest, usually a little shorter than the following· ones, broadest part of the laminilusually just below the middle. Pinnae usually symmetrical (with acroscopic and basiscopic segments £ the same length) and

2 The following abbreviations are used: =· for homotypic synonyms, = for heterotypic synonyms, - for misapplied names (see below). FRASER-JENKINS, REICHSTEIN, VIDA: DRYOPTERIS TY RRHENA NOM. NOV. 181

divided in the manner of D. filix-mas, i.e. only the basal pair of segmentscompletely separated from the next. Basal segments .:t. oval, often slightly auriculated but not or only scarcely petiolate; further segments similar but sessile and contracted at the base which becomes increasingly broader and decurrent higher up the · pinna. Extreme forms of very big fronds have the largest basiscopic pinnules of the lowest pa ir of pinnae deeply dissect (as in D. villarii subsp. pallida) and c. 1.5-2 times as long as the acroscopic, thus making these lowest pinnae strongly asymmetric (fig 8). All pinna segments (except the very smal l ones) are usually more distinctly separated from each other and less fused at the base than in normal forms of D. filix-mas, their apex is ± circular, and finaly serrate with acute teeth, separated c. 0.5-1 mm from each other and tending to splay out after the manner of Q. abbreviate but often curving inward at the tips (figs 3 and 6). the margins are more coarsely and partly twice dentate. Sori c. 1 mm diameter, 1-12 per segment crowded at the base of the segments, leaving the upper 1/3-1/2 bare. Indusia similar to those of D. abbreviate but si ightly 1·ess convex and covered with glands over the whole surface, becoming lifted at the margin when the sporangia are ripening. Whole lamina on both sides including rach is, densely covered with shortly stalked glands, easily visible FIGURE 2. Aspidium nevadense Boiss. with a lens even after pressing. Spores bean Original specimen in G. shaped with distinct ridges similar to D. villarii in appearance (32-) 4Q-42 (-;-46)p m long (fig 10). Tetraploid (2n = 164), sexual,with 64 �ores per sporangium.

TA BLE 1. Some characters which usefully distinguish the four critical Dryop teris species, not including extreme forms.

D. villarii D. tyrrhena D. filix-mas D. abbreviata subsp. 1 2 3 4 pallida

Length of ripe fronds, in cm (1Q-) 3Q-70 1Q-50 (-60) 3Q-80 (- 1 20) 20-70 (-90) Relative length not or· usually a usually tapering of pinnae scarcely little distinctly markedly versus base reduced reduced reduced often from as as far up as the the middle Longest pairs of pinnae (counted from number number number number base) 1-5 (1-) 3-6 (4-) 6- 14 (6-- )8-16 Widest part- Much below Just a little Just a I ittle Distinctly above of lamina the middle below the above the the middle middle middle

(footnotes 1-4, see p 183) continued on following page 182 FERN GAZETTE: VOLUME 11 PART 2 & 3 (19715)

D. villarii D. tyrrhena D. filix-mas D. abbreviate subsp. 1 2 3 4 pallida

Base of the petiolate, The lowest The lowest pair The lowest pair larger pinna usually with pair separate separate from separate from segments (on a stalk of from the next the next but the next but lowest part 1-2 mm but not petio- not petiolate not petiolate of pinnae) late, the next the next pairs the next pairs pairs sessile sessile ± fused not fused but not fused Apax of the narrow ± ±.circular usually narrow usually round pinna pointed with with acute but not acute, ±circular with segments acute teeth teeth often sometimes rou nd obtu se-tipped curving inward with acute teeth spread ing teeth Margins of the coarsely and partly usually usually pinna deeply lobate, distinctly shallowly lobate shallowly segments lobes with lobate, lobes with acute lobate, with acute teeth with acute teeth obtu se teeth teeth

Sari: diameter(mml c.1 (-1.5) c. 1 (- 1.51 c. 1 . 2-2 c. 1 (-1.5) Amount of nearly to· c. 1/2-2/3 c.1/2-2/3 c.1/2-2/3 segment apex (-3/4) of (-3/4) of (-3/41 of covered by basel part basal part basal part sari Indusium convex at first, convex at first, usually convex, glabrous ± flat later, ±flat later, ± flat, though except for glands strongly strongly somewhat convex on the margin glandular glandular when growing on occasionally. dry or exposed conditions, glabrous Av. length of (28-) 30-34 (35-) 38-42 (36-) 40-46 (30-) 32-36 5 spores in,um (-40) Glendulosity both sides both sides both sides usually ± of lamina and usually strong very strong ±glabrous glandular rhachis particularly on the lower side, less so in Northern Europe

Scales on straw coloured straw coloured yellowish or reddish stipe or light or light redd ish brown brow n reddish brown reddish brown

Persistence of always nearly always only in rare never fronds through cases winter Favoured limestone N-exposed humus, woods N-exp. silicate habitats rqcks, walls silicate rocks, but also in rocks, rocks, boulders, screes, sunny c·avE!S, rocks, walls screesand screes. near ± bushy places crevices. edged of road-sides water courses, screes sometimes road-sides Favoured 0-800 m 0-2000 m 0-2000 m 800-2000 m altitudes

Ploidy 2n = 82 2n = 164 2n = 164 2n = 82

(footnote5, see p 183) FRASER-JENKINS, REICI-61EIN, V lOA: ORYOPTERIS TYRRHENA NOM. NOV. 183

Footnotes to Ta ble 1.

1 Common on Sardinia, absent or very rare on Corsica (see below). 2 Rather rare everywhere. 3 Rare o�Sardinia more frequent On Corsica. 4 Often growing in big colonies on the higher mountains of Sardinia, Mt Gennargentu and Mt Limbara (CRFJ & TR 1974) but more frequent on Corsica. 5 Exospore measured without perispore, in balsam. Only good spores measured and not stunted or damaged ones. Contents of complete sporangia exam ined, having been collected from fronds that are just ripe, not residues from herbarium specimens.

INTERPRETATION AND TREATMENT OF DRYOPTER/S TYRRHENA BY FORMER AUTHORS The identity of Aspidium filix-mas Sw. var. glaridulosum Milde as Dryopteris abbreviata D. tyrrhena is often fo und in many herbaria under names based on Aspidium filix-mas var. g/andulosum Milde. lt was therefore advisable to examine Milde's type even if for our purpose {to select a valid name for Aspidium nevadense) it was not · strictly necessary. Milde quotes for his var. glandu/osum the followi ng two specimens "In Monte .Gennargentu Siciliae3 . (A scherson) . Corsica {Requien )". We hesitate to select one as lectotype as the identity of the specimens is not completely certain. We do not know for certain whether Milde's first specimen still exists tllough it would seem to be identical with the specimen of Ascherson & Reinhardt in {B): "Steinige Abhange des Xuxu, Gennargentu Sardoa 1863" determined by Milde asvar .glandulosum and mentioned. . in Fiori (1943: 100, translated into cm 3 italian). CRFJ identified this specimen in {B) as D. abbreviata. He also saw three specimens of the second ongm: "Requien, �ciut 1847 For�t d'Aitone" (see Briquet 1910: 10) deposited in {B), (P) and {G) and idEmtified all three as D. abbreviata. This species is rather common in the For6t d'Aitone but in spite of careful search we (CRFJ & TR, July 1974) could not find D. tyrrhena there. Further arguments supporting the synohlyny of Aspidium · filix-mas var. glandu/osum with Dryopteris abbreviata are: 1. Milde's description of var. glan­ dulosum fits D. abbreviata much better than D. tyrrhena (except for glandular indusia, though D. abbreviata often has minute glands on the margins of the indusia). 2. We do not believe that such a careful worker as Milde would publish one and FIGURE 3. Aspidium nwadflnsfl Boiss. Single pinnae of original specimen (G) cor­ 3 This must be an error because Mt responding to fig 2. (Ph oto. L. Jen ny). Gennargentu is in Sardinia, not in Sicily. 184 FERN GAZETTE: VOLUME 11 PART 2 & 3 (1975)

E u CJ ..-- 0 I I

Q)

ro'

FIGURE 4. Silhouettes of fronds x c. 1/5. a, D. abbreviata 3558; b, D. tyrrh ena 3562; c, D. tyrrhena hybrid 3789 (4); d, D. villarii 4x 3568; e, D. x sardoa 3775; f, exp. hybr. GV579. FRAS ER-JENKINS, REICHSTEIN, VIDA: DRYOPTERIS TYRRHENA NOM. NOV. 185

the same taxon (D. tyrrhena) under two different names (A sp dium rigidum forma pinnatisecta and A. filix-mas var. glandulosum) in the same pub I ication. 3. D. abbrevia ta is much more frequent on Mt Gennargentu (Sardinia) than D. 4 tyrrh ena . We are therefore confident, that M ilde's A. filix-mas var. glandulosum is indeed what we call today D. abbreviata and that 1\11 ilde did intend and was able to separate it well from Aspidium nevadense Boiss.

The identity of Dryopteris litardierei Rothmaler pro hybr. Under this name we have seen only one specimen of D. tyrrhena, but it was an important one and we therefore discuss this taxon: Dryop teris litardierei Rothm. was originally interpreted a!i a hybrid D. pa/eacea x pallida by Rothmaler (1945 : 94-95), who c-ites: "Korsika, Evisa, Foret d'Aitone, 1 '1 00 m. Sagorski = Typus". This specimen (JE) was identified by CRFJ as [). abbrevi­ ata. The name D. x litardierei is therefore a synonym of the latter species. Roth m aler equates his putative hybrid with Milde's var. glandulosum and states that he has seen Milde's type. The three following specimens (in 8), identified by Rothmaler as D. x /itardierei, seen by CRFJ, were all. D. abbreviata: 1. Foret d'Aitone, Corsica (Requien), det. by Milde (1867: 123) as Aspidium filix-mas var. glandulosum; 2. Steinige Abh�nge des Xuxu, Gennargentu, Sardoa (A scherson & Reinhardt 1863) det. by Milde as var. glandulosum; 3.

4 The authors saw (July 1974) a big colony of D. abbrevia ta (more than 100 plants) on the N side of Bruncu Spina, beside a stream W of the new road at c. 1500 FIGURE· 5. Silhouette of frond of D. villarii m alt. and only c. 0.5 km N of the lower subsp. pal/ida CRFJ 2126, x c. 5-· sk i-1 ift station. 186 FERN GAZETTE: VOLUME 11 PART 2 & 3 (1975)

Mte. Renoso, Corsica (Requien). But another specimen:. Liguria. Riomaggiore, leg. Bornmuller (B), designated by Rothmaler as "D. paleacea x pallida Rothm. (Hybr. nova) Typus! (Original)" was identified by CRFJ as D. tyrrhena. Had Rothmaler selected this specimen as type in his publication (as he probably originally intended to do ) we would have to use the name D. litarierei Rothm. for Aspidium nevadense Boiss.

Other attemptsto distinguish the critical taxa As mentioned above, Milde was able to distinguish his Aspidium filix-mas var. glandu/osum (=D. abbreviata ) from A. rigidum "forma pinnatisecta" (=D. tyrrhena ). Some, but not all of his successors were able to do the same, to mention Litardiere, Briquet, Rouy and Fiori. l.itardi?ne (partly together with v. Tavel) has in his later publications (1924, 1928 and the three following not dated) probably distinguished "var. nevadensis" from "var. glandulosa'�. Of the ma ny specimens he quoted (1924) for "var. nevadensis" we have so far examined only the 4 mentioned below (sub Corsica), all correct D. tyrrhena. Those he lists under "var. glandulosa" are probably all D. abbrevia ta. CRFJ examined several correctly determined collections of 1908 (P, CLF & BR) and the following, quoted in Litadiere (1924: 125): "Corse, Massif du Rotondo: foret de Polverella, ravin de la Tasseta. N.W. de Corte, 1,070-1.120 m. (Litad/ere 31. aout 1919)" (P), which was again correct D. abbreviata. The bulk of Litadiere's he.rbarium, however, is owned by his family and has not been seen by us. Briquet (1910: 10) does not mention Aspidium nevadense, but gives sub D. filix-mas var. glandulosa (M ilde) Briq. the specimen of D. tyrrhena mentioned below (see p. 189) and eight other specimens, of which the five following have been seen by CRFJ and were all identified by him as D. abbreviata : Requien 1847 Fo�et d'Aitone, Gorse (G, B & P); Lit. (see Litardiere 1908) Monte Grosso (C LF) Mt. d'Oro (P& BR); Briquet 1906 Mt. Rotondo, rocher a 2600 m, 6 aout (G); and 1908 Monte Asto, creux des roches, 1500 m. 1. juill. (G). These other localities given are: Lit. Vallon de Taita; Lac du Capo Felo; and Lago Maggiore sous le Capo al Berdate. However although these are mostly D. abbreviata, from notes written by Briquet on specimens in Geneva, it seems that D. tyrrhena was the nearest to his concept of var. glandulosa, being more glandular. �ouy (1913: 408-409 ) does distinguish, theoretically, between Aspidium nevadense Boiss. and his Nephrodium filix-mas race rigidoformis (based on A. filix-mas var. glandulosum Milde) but quotes 4 specimens under the second name, of which the first (Requien) is D. abbreviata, the next (Foret 'Asco) real D. tyrrhena, and the last two (Rotondo & .Asto ) again D. abbreviata. These specimens were determined by CRFJ at (G). Fiori (1943: 100 & 113- 114) distinguishes between D. filix -mas var. glandulosa and D. villarii var. nevadensis. One of the specimens he quotes sub var. glandulosa for Corsica (Foresta d' Asco) is D. tyrrhena, but most of the other ones are D. abbreviata, except "Piem.-Aipi Mar. fra Trappa e Garessio e pr. Ormea a Chioraira (Burn. ex Christ). CRFJ could find Burnat's specimens (in P), they were 3 D. borreri (with glands). Chioraira was also visited in June 1974 by H.L. and T. Reichstein, D. filix-max and D. borreri with hybrids are abundant l:iut no other Dryopteris was seen. Among the 7 specimens which Fiori (1943: 114)' quotes for var. nevadensis only those from Elba and Capraia are correct 0. tyrrhena; we have not seen "Is. d'lschia (Bolle · ex Beg.)", but "Laz. al M. Circeo, tra S. Felice ed il Faro (Somm. ); "Sa rd . tra Ulassai e la V. del Tarquisara (Somm., det Christ)" and "V. del Tarquisara (Biondi)", (all in F I) are D. villarii, most probably subsp. pallida. We have visited the two last FRAS ER-JENKINS, REICHSTEIN, VIDA: ORYOPTERIS TY RRHENA NOM. NOV. 187 mentioned localities (C RFJ & TR in July 1974). The altitude is from c. 740-800 m with limestone- rocks. D. vil/arii is abundant, probably as subsp. pallida (but we cannot exclude some tetraploids, th is wi ll be checked when our living plants grown from spores can be counted). Among plants from these localities were a few which produced single fronds simulating D. tyrrhena rather well. Without careful control such fronds can easily be confused. But additional fronds from the same plant usually make identification much easier.

DISTRIBUTION OF D. TYRRHENA We have unequivocal material of D. tyrrhena, including living plants or viable spores collected by ourselves or obtained from friends and have seen specimens in different herbaria, from the fo llowing places:

Southern Spain Besides the three type specimens (in G) we have seen the following : N-exposed silicate rocks above scree on southern side of upper Dilar valley, Western Sierra Nevada, at c. 2300 m alt. 6 Aug. 1974 leg Fraser-Jenkins 4369 (BM). The species has already been collected in the same area by Spanish botanists, e.g. by Clemente "Bajolas penas er� las Lagunillas de Sierra Nevada" (MA). When identifying herbarium specimens it is important to remember that D. filix -mas and D. villarii (the tetraploid taxon ) also grow on the Sierra Nevada the latter at slightly lower elevations and exclusively on limestone. Both D. tyrrhena and D. vi/larii are rare there (see below).

Sardinia Massif of Mt Gennargentu: Mt Spada, silicate rocks on E face but near the NW ridge, mostly in crevicess or under overhanging rocks between c. 1400-1 500 m alt. leg. Fraser-Jenkins 43 10 & Reichstein 3776, 25 July 1974; sporadic single plants or small groups, sometimes together with a few D. abbreviata and one hybrid (see below). Sardinia, no exact local ity, leg. J. Bornmul/er (B; det. as var. glandulosa by Rothmaler). This is D. tyrrhena.

Corsica 1. La Trinite, WNW of Bonifacio: between granite boulders in a cave below a slight overhang at the base of the N facing cliff of the western peak at c. 150 malt., huge, obviously very old colony, leg H. L. & T. Reichstein T�-3088, 16 June 1970. We found the place with a description and sketch obtained from J. Prudhomme, who had sent us in July 1967 a frond with spores from which progeny, TR-2047, were raised in Basel for cytological contro,l . lt was tetraploid, 2n = c. 164 (G.V. 1 June 1969; fig 11 ). Specimens of this plant (in BR, G and P) have been collected by Stefani 1912 at the same spot. The place was rediscovered independently by J. Vivant on 28 March 1967 who informed Prudhomme. D. tyrrhena (CRFJ 4345; TR-3789-5, 6, 10, 11, 13) grows there with fronds up to 60 cm long, some with strongly asymmetric basal pinnae (fig 8) together with a hybrid producing fronds up to 85 cm long (see 3789-4 in fig 4) and abortive spores, leg. Fraser-Jenkins CRFJ-4346 and Reichstein TR-3789- 1, 2, 3, 4, 7, 8, 9, 12, 31 July 1974. This hybrid may be D. filix-mas x D. tyrrhena and is being examined further. 2. Calanche de Piana: N-exposed scree with granite boulders c. 50 m below the road N-199 from Porto to Piana, c. 350 m west of the Auberge: "Les Roches Bleus" 188 FERN GAZETTE: VOLUME 11 PART 2 & 3 (1975)

2.cm

1

a

0

FIGURE 6. Pinnae from ± widest part of fronds cleared in c. 70% chloral hydrate to show venation; indusia are lost in this procedure. x 2, drawn by R. H irzel. a, D. filix -mas,· b, D. tyrrhena (CRFJ-4350), the inward curving of the segment teeth is somewhat exaggerated (this character is not al�ays present but if visible is highly significant and only rarely present in a and c), detai l x 4; c, D. abbreviata (3628). FRASER-JENKINS, REICHSTEIN, VIDA: DRYOPTERIS TY RRHENA NOM. NOV. 189 c. 400 m. alt. 3 colonies, leg. H. L. & T. Reichstein TR-3562, 26 June 1973 following indications of H. & K. Rasbach who found th is place, 4 May 1971 (see fig 1 ). Some other good plants were also found c. 80 m higher, i.e. above and south of the road in the rocks at c. 480 m alt. leg. Fraser-Jenkins CRFJ-4350 and Reichstein TR-38 12, 1 Aug. 1974. 3. Mt Renoso, pr'es sommet. leg. R. Descha tres 1967. 4. Mt Renoso, Rocher pres d'une pozzine sous le lac de Bastani al. 2050 m env. leg. R. Deschatres 8. aout 1974. 5. South of Mt Cinto: Silicate rocks at the junction of the rivers Viro and Golo ("rives droites de chaque rivi'ere"), c. 1 km SW of Albertacce, c. 820 m alt. leg. R. Deschatres 21 juillet et 9 ao{J't 1974; together with D. abbreviata and hybrids (see below). 6. Massif du Cinto (bass in sup. de I' Asco): Rochers sur la rive gauche de I'Asco pres de la Resinerie de la foret d'Asco 1200 m, leg. Burnat, Briquet, St. Yves, Cavillier et Abrezo/, 20 juill. 1906 (G) sub D. fillix-mas var. glandulosa. (Briquet 1910: 1 (}) , 7. Asco, Rochers au Pinzalone, au-dessous des bergeries d'Entrata, 1050 m env., Litardil!re 28 juill. 1921 (CLF) sub D. filix-mas subsp. rigida var. nevadensis (Litardiere, 1924: 124). 8. Asco, Vallee de Violini, rochers 1100 m env. Litard/ere 25 juill. 1921 (CLF) sub same name. 9. Environs d' Asco, au Fornello di Grosso, rochers porphyriques, 1060 m env., Litardiere, 22 juill. 1921 (P) sub D. filix-mas subsp. rigida var. nevadensis forma ro tunda ta Lit. et Tavel, (Litardiere, 1924: 124 et figs 2 & 4). 10. Asco, route de Stagno, For�t de la Carozzica. leg. J. Calle & R. Deschatres, 10aout 1967. 11. Cap Corse: Southern slope of Mt Stello, near the small path from Pozzo to the Bocca di Santa Maria at the little shepherd 's hut above a spring, c. 900 m. alt. between silicate rocks and boulders, localy abundant, leg. H. L. & T. Reichstein TR-3551, :n June 1973 now cult. in Basel. Tetraploid n =c. 80 (G.V. 5.2.1974). We visited this place because of a specimen of Jaquet 1921 sub. D. filix-mas var. glandulosa (Milde) Briq. in (G) which was D. tyrrhena. 12. Cap Corse: Mt Canneto, versant SE rocher 1220 m env. Litardi'ere 17 juillet 1921 (P) sub D. filix-mas subsp. rigida var. nevadensis (Litardiere, 1924: 124).

Italy, mainland Liguria, Riomaggiore (W of La Spezia) leg. Bornm'JIIer. Spec·imen (B) designated by Rothmaler as "D. paleacea x pal/ida Rothm. (Hybr. nova) Typus! (Original)", discussed above.

Capraia Fiori (1943 : 113-1 14) gives for his D. villarii var. nevadensis apart from Elba (see be low) the Is. Capraia, a Cala della Mortala. This specimen, Sommier 1896 sub. Aspid. filix-mas ( F I) is D. tyrrhena.

Elba Vineyard walls (silicate ) and transverse ditch c. 50 cm deep, near R ipa Baretta 'W of Mardana Marina c. 110 m alt. 6 plants with fronds 13-45 cm long all fertile leg. F. Mokry, 24 May 1974. Spores sown in Basel (TR-3130). Fiori (1943 : 113-1 14) gives 190 FERN GAZETTE: VOLUME 11 PART 2 & 3 (1975) this place correctly under his Dryop teris villarii var. nevadensis mentioning a specimen of Sommier (1900) sub. Aspidium fi/-mas var. glandu/osum� Th is specimen (in Fl) has been confirmed as D. tyrrh ena. Obviously this list does not claim to be complete and we suspect that other specimens of D. tyrrhena may be found in different herbaria under different names. On the other hand other species, particularly D. villarii, D. abbreviata and D. filix -mas are often deposited in herbaria under some of the names used for D. tyrrhena.

DR YOPTERIS VILLAR/1 ON·CORSICA As D. tyrrhena can be confused with abnormal forms of D. villarii and also when considering its possible parentage we were interested to know the distribution of D. vt11arii on Corsica. The result was unexpected. Briquet (1910: 11-12) gives for his D. rigida Underw. var. meridio nalis Briq. three localities: 1. Mt Cinto 2000-24000 m. (Briq. Rech. Corse 107 et exsicc. Burn. ann. 1900, n. 129); 2. col de Tavoria, 1600 m (Rotges in litt.); 3. Pointe de la Monte, au-dessus du col de Verde, vernaies du versant W, 1600-1 700 m (20 juill. 1906). We were able to see the first and th ird specimens in (G) (C RFJ-1973). these being D. abbreviata: The two last mentioned places were also visited by H. & K. Rasbach and H.L. & T. Reichstein on 26 and 27 July 1973. In both places D. abbreviata was growing abundantly. A living specimen (TR-3624) was collected at the Col de Taoria (c. 1600 m) and turned out to be diploid (2n =c. 80 det. G.V.). Both places consisted of granitic scree and boulders, the second with Alnus viridis, completely unlikely habitats for D. villarii. CRFihas also seen specimens (in P) of Rotg'e s "Corse, Ghisoni, Col de Tavoria, 1400 m, 1 Sept. 1899", sub Polystichum rigidum, which were again D. abbreviata.

For his var. austra lis Briq. (= subsp. pallida l he also gives three places: 1. Ch1ltaigneraies pres de Bocognano, 600 m (Briq. ),· 2. foret d'Ospedale (Seraf. ap. Viv. I.e.) and 3. Cime de la Chapelle de S. Angelo, rochers et balmes, 1100-1180 m. calc. 15 juill. 1906 et 13 mai 1907. The specimen from Bocognano is in (G) and is D. abbreviata. In summer 1970 CRFJ visited the place and found a few plants of D. abbreviAta, an extremely low altitude for th is species in Corsica. In the foret d'Ospedale we searched in vain for D. villarii, the rocks there are silicate and the soil acid, it is extremely unlikely that D. villarii would grow there. We have not yet fou nd the original specimen (Seraf. ). The only unequivocal specimens of D. villarii from Corsica we have seen (in G examined by CRFJ) came from the Cime de la Chapelle de S. Angelo. They are either subsp. pal!ida or the tetraploid taxon. A final identification is difficult without cytological control and curiously the plant has not so far been found again. The place is given with great precision and cannot be missed. The whole area (1 100-1 180 m) which corresponds with the top of the hill contains many shallow and deep cracks in the white karstic limestone pavement and is exposed in all directions. lt is not a very large area and is an ideal habitat for D. villarii. The place was searched independently and rather carefully by Fraser-Jenkins on 13 July 1970 and again on 29 June 1973 by H.L. & T. Reichstein. No trace of D. villarii could be found. R. Deschatres, who has an excellent knowledge of the Island, told us that he had the same experience and also that he had never seen D. villarii in any other place in Corsica. The specimens in (G) show, however, that it must once have grown there but as a great rarity. In Sardinia, on the other hand, the species is rather common in most of the limestone areas (e.g. de Joncheere 1963). we have the diploid subsp. pa!lida from the east coast (TR- 117, FRASER-JENKINS, REICHSTEIN, V lOA: ORYOPTERIS TYRRHENA NOM. NOV. 191

FIGURE 7. Base and apex of fertile pinna of D. villarii subsp.pa llida TR-3588, cleared as in fig 5.

1.1 0-1959; n = 41 det. G. V.) but are not yet sure whether the tetraploid also grows on the island ; the alpine diploid is very unlikely to be there and we have not seen any specimens of it from Sardinia.

POSSIBLE ORIGIN OF D. TYRRHENA As pointed out in the introduction D. tyrrhena is intermediate in morphology between the members of the D. filix-mas and D. villarii groups and we suspect that it once arose from a diploid hybrid between a representative of each of these groups, with subsequent doubling of its chromosomes. Provided that the putative ancestors are still living and grow in Europe or not far from it, the most suitable candidates would be D. abbreviata and D. villarii subsp. pal/ida. Both are dipioids and both are distinctly glandular albeit not usually as pronounced as in D. - tyrrhena. But it is known that certain characters may be not only suppressed but sometimes enhanced by hybridisation. We suspect that subsp. pallida rather than the alpine dipl. subsp. villarii is the second ancestol' because in D. tyrrhena the frond tapers only si ightly towards the base, or sometimes not at all. The chances of fi nding a diploid hybrid of D. abbreviata x D. villarii subsp. pal/ida in nature are not very great Qecause normally these species grow in quite different habitats. But it could be formed under special conditions, as when granite or schists and limestone meet near a water course and we have already seen single plants of D. pallida on granite in rare cases. Under present conditions it cou ld perhap� form on Sardinia but not on Corsica. Provided our hypothesis is correct it would be another example of how an alpine species (D. abbrevia ta ) and a lowland one (D.· villarii subsp. pallida l form an allopolyploid plant (D. tyrrhena) which grows from the lowland to high alpine level. An other example which is known is Polystichum aculeatum (L.) Roth. (growing from lowland to alpine level) which according to Manton (1950, see also Manton & Reichstein 1961 ) has arisen by chromosome doubling from a hybrid of the alpine P. lonchitis (L.) Roth., and the lowland P. setiferum (Forskal) Woynar. The relict type of distribution (rare and scattered with big gaps between different localities, partly in caves) suggests that D. tyrrhena is an old allotetraploid, perha-p s formed at a time when 192 FERN GAZETTE: VOLUME 11 PART 2 & 3 (1975)

FIGURE 8. Lowest pinnae, cleared as in fig 5. a, D. villarii subsp. pallida (3588); b, extreme form . of D. tyrrhena, strongly asymmetric (3789-10). · FRASER-JENKINS, REICHSTE IN, VIDA: DRYOPTERIS TY RRHENA NOM . NOV. 193 the putative parents were slightly different from the present representatives of the me ntioned species. To test this hypothesis we intend to examine the cytology of suitable hybrids. particularly their behaviour in meiosis. These experiments have only been recently started and need much time but we already have two important hybrids, a natural and an artificial one, which re present exactly the putative backcrosses which we need.

HYBRIDS Dryopteris x sardoa Fraser-Jenkins & Reichstein, hybr. nova (D. abbreviata x D. tyrrhena) Diagnosis: Planta hybrida media inter parentes praesumptos D. abbreviatam et D. tyrrhenam sed illi similior. Frondes validae ad 55 cm longae, eis D. abbreviatae similes. Ab hac disti nguitur: 1. lamina ad basim versus minus angustata; 2. squamis latioribus (ad 8 mm); 3. sporis abortivis; 4. numero chromosomatum somaticorum c. 123. Type: Sa rdinia, Massif of Mte Gennargentu, in crevices of the silicate rocks at the east face of Mt Spada near the NW ridge at c. 1500 m alt., close to a plant of D. tyrrhena and with a plant of D. abbreviata c. 10 m away, 25 July 1974. Fraser-Jenkins & Reichstein 3175. Holotype G; lsotypes B, BM, F I, K, P (fig 4). We took 2 pieces of rhizome, one (TR-3775 ) is cultivated in Basel and one (CRFJ-4309 ) in Abingdon. Two or three plants of the same hybrid have been found by R. Deschatres in Corsica: south of Mt Cinto at the junction of the rivers Viro and Golo c. 1 km SW of Albertacce, c. 820 m alt. between the parents, 21 July and 9 August 1974. The hybrid has characters intermediate between the two putative parents D. abbreviata and D. tyrrhena but is more similar to the former. The fronds are no more glandular than in D. abbreviata but in outline taper slightly less towards the· base, the scales at the base of the stipe are a little wider (up to 8 mm, fig 9). For final determination it is essential to ex amine the content of ripe sporangia showing abortive

!ern

0

3628 3588 O.abbr. D. pa ll. D. xsard .

Fl GUR E 9. Scales from petiole bases. Left to right: D. abbreviata; D. tyrrhena; D. villarii subsp. pallida; D. x sardoa. (Ph oto. L. Je nny). 194 FERN GAZETTE: VOLUM'E 11 PART 2 & 3 (1975)

..... 748 ,::� 3623 D.fit.-mas a D. abbr. b - D. ty rrh .

c

3775 e �\�·

FIGURE 10. Contents of ripe sporangia x 200: a, D. abbreviata; b, D. filix-mas; c, D. villarii su bsp. pa/fida; d, D. tyrrhena; e, D. vil/arii subsp. vi/farii,· f, D. x sardoa. (Ph oto. L. Jenny). FRASE R-JENKINS, REICHSTEIN, VIDA: DRYOPTERIS TY RRHENA NOM. NOV. 195

FIGURE 11: Mitosis of D. tyrrhena, showing c. 164 ch romosomes, x c. 1500.

spores (fig 10). So far only mitosis has been examined. Th is shows that D. x sardoa is a triploid hybrid, as expected · with c. 123 chromosomes. Further results will be pubI ished later.

Experimental hybrid GV-579 of Dryopteris ty rrhena x D. villarii subsp. pallida This was produced by V ida in Budapest using methods similar to those described earlier by Lovis (1968). Prothalli of D. tyrrhena (TR-3088) were used as � and of D. villarii subsp. pallida (TR-949) as cf.The latter was raised in Base I from spores collected by E. H auser on I imestone rocks on Mt Bondone (N Italy, near Trento) c. 400-700 m alt. where it is rare; diploid (n = 41, det. G.V. 17 June 1969). Hybridisation was done on the 12 March 1971; 23 Q prothalli gave 2 hybrids. Fig 4 shows a frond and fig 12 the contents of ripe sporangia. Cytological, examination showed that the plant is a triploid hybrid producing at

··' FIGURE 12. Sporangia! content (abortive spores) of ex per. hybrid GV-579, x c. 200. _(Ph oto. L. Jenny). 196 FERN GAZETTE: VOLUME 11 PART 2 & 3 (1975) meiosis c. 40 bivalents, the other chromosomes were present as univalents. Details will be given later.

CONCLUSION AND DISCUSSION The above preliminary cytological results are compatible with the assumption that Dryopteris tyrrhena is an allotetraploid species formed originally from a diploid hybrid of D. abbreviata (genomic formula AA) with D. villarii subsp. pallida (PP) and therefore should possess the genomic formula .AAPP. The back cross with D. abbreviata should be a triploid hybrid with thege nomic formula AAP and should show at meiosis c. 41 pairs and c. 41 univalent chromosomes. Whether or not D. x sardoa will show this behaviour we hope to ascertain next summer. But the fact that this hybrid was found twice in nature at places where both parents grow together but are represented only by few plants makes it probable that this hybrid is formed easily. In the genus Dryopteris ready hybridisation has so far been observed only in those cases in which both parents contain a common genome (e.g. D. abbreviata x D. filix-mas; D. assimilis x D. dila ta ta; D. carthusiana x D. dilatata etc.). The relationship between D. borreri and D. filix-mas, which also hybridise readily, is not quite clear. In a similar way the experimental hybrid of D. tyrrhena xpallida which should possess the formula APP, also shows the expected behaviour (c. 40 pairs). The proof in D. tyrrhena sti ll missing so far that · is an allo- and not an autotetraploid. Other hybrids will be necessary to prove this. D. tyrrhena (sub. D. filix-mas var. rigidoformis) has also been subjected to chemical analysis (Widen et al. 1971 ). lt contains in its rhizomes seven phloroglucides of which two, albaspidin and trispara-aspidin (together with traces of two others) are not present in D. abbreviata nor in D. fil ix-mas or D. caucasica, (see Widen et al. 1973) white desaspidin was absent in D. tyrrhena and in D. abbreviata, but present in D. filix -mas and in higher concentration in D. caucasica. These results fit very well with the fact that albaspidin and trispara-aspidin are found in relatively high concentration in D. pallida while only little desaspidin is present. This would be expected if the phlorogtucides of D. tyrrhena are ± the sum of those present in its putative parents provided we assume that formation of para-aspidin (present in high concentration in subsp. pallida but lacking in D. abbreviata and D. tyrrhena ) is inhibited in D. tyrrhena by the D. abbreviata genomes. Although chemical analysis can only be regarded as supplying one more character, these results are highly compatible with the assumption that D. tyrrhena contains two genomes each of D. abbreviata and D. villarii subsp. pallida and they also correlate well with the assumption that D. filix -mas contains two genomes each of D. abbreviata and D. caucasica. Final proof must come from cytological analysis of all the necessary hybrids.

Acknowledgements We would like to thank the directors and keepers of th e following herbaria for th eir help in examining specimens: B, BM, BR, Fl, G, JE, K, P and Messrs. F. Badre (Paris). R. Deschatres (Escu rolles). F. Mokry (Au, Switzerland). Dr. E. Nard i (Firenze). J. Prudhomme (Neuville-sur-Sa�ne), J. Vivant (Orthez) for pressed fronds and most valuable indications of localities, sketches and maps for Corsi_c a and Sardinia; Mrs H. and Dr K. Rasbach (Giotterbad, Germ any) and Professor L. Jenny (Basel) for photographs and Mrs R. Hirzel for line drawings; Dr J. v. Euw for silhouettes of fronds; Dr W. Greuter (Geneve) and A.C. Jermy (London) for advice in nomenclature questions and their help in preparing· the manuscript, and Dr H. HUrl imann (Basel) for Iatin diagnosis. FRAS ER-JENKINS, REICHSTEIN, V lOA: DRYOPTE RIS TY RRHENA NOM. NOV. 197

REFERENCES ASCHERSON, P. & GRAEBNER, P. 1896; 1910. Synopsis der Mitteleurop'Jischen Flora I. ed .. l. et ed. 2. Leipzig. BAKER, J.G. ( 1891 ) . A Summary of new Ferns discovered since 1874. Ann. Bot. 5, 320. BOISSIER, E. 1838. Elenchus plan tarum novarum minusque cognitarum, quas in itinere hispanico legit. (lenevae. , ,.. BRIQUET, J. 1910. Prodrome de la Flare Gorse I. Geneve et Bale. CHRISTE NSEN, C. 1906. Index Filicum, Hafniae. COPELAND, E.B. 1947. Genera Filicum New York and Waltham, Mass. EATON, A. C. 1878. The Ferns of North America. Coloured Figures and Descriptions of the Ferns (including the Ophioglossaceae) of the United States of America and the British North American Possessions. 1 ( 1877), 73, t. 10. Sal em. FABERGE, A.C. 1945. Snail stomach cytase, a new reagent for plant cytology. Stain Te chnology 20: 1-4. FIO RI, A. 1943. Flora ltalica Cryp togama. Pars V: Pteridophyta. F irenze. FRASER-JENKINS, C.R. & JERMY, A.C. (in press). Nomenclatural notes on Dryopteris Adans. Taxon. . FRASER-JENKINS, C.R. & CORLEY, H.V. 1972. Dryopteris caucasica - !lnancestral diploid in the male fern aggregate. Brit. Fern Gaz. 10: 221-231. GILBERT, O.L. 1966. Dryopteris villarii in Britain. Brit. Fern Gaz. 9: 263-268. HEYWOOD, V.H. (1964) in Flora Europaea I, 21-22. Cambridge. HOLUB, J. 1967. Remarks on the Nomenclature, of "Dryopteris borreri Newman 1854". Folia geobot. et phytotaxon. 2: 329-332. : JONCHEERE, G.J. de. 1963. Ferns of Sardinia. Brit Fern Gaz. 9: 114- 116. LITARDI�RE, R. de. 1907. �oyage .botanique en Corse. Bull. Soc. Bot. Deux-Sevres 18: 125-150. LITARDIE RE, R. de 1908. Voyage botanique en Corse (1907). Bull. Acad. Geogr. Bot 19: 135- 1 68. (Contains only few indication about ferns). LITARDIE RE, R. de. 1911. Contribution 'a l'etude de la Flore pteridologique de la peninsula iberique. Bull. Geogr. Bot. 21: 12-30. LITARDIE RE, R. de. 1914. Contribution a l'�tude de la flare de Corse. Bull. G(ogr. Bot. 24: 8� 108. LITARDIERE, R. de. 1924. Contributions ·a l'�tude de la Flare de la Corse. Notes sur quelques Filicinees du Cap, des massifs du Cinto et du San Pietro. Ann. Soc. Linn. Lyen, N.S. 70: 12.1-133. LITARDIE RE� R. de. 1928- 1930. Nouvelles contributions ·a l'etude de la Flare de Corse. Arch. Bot. 2: 1-44, 3: 1-30, 4: 1-16, 5: 1-10. LITARDIE RE, R. de & MA LCUIT, G. 1926. Contributions � /'etude phytosociologique de la Gorse. Le massif du Renoso. Paris. LOVIS, J.D. 1968. Fern hybridists and fern hybridizing 11 Fern hybridizing at the University of Leeds. Brit. Fern Gaz. 10: 13-20. LU E RSSEN, C. 1889. Die Farnpflanzen oder Gef�sskryptogamen (Pteridophyta) in L. Rabenhorst Kryptogarnen-Fiora von Deutschland, Oesterreich und der Schweiz. 2. edit. 3 Leipzig. MANTON, I. 1950. Problems of cytology and evolution in the Pteridophyta: Cambridge. MANTON, I. 1961 . Evolution in the Pteridophyta in P.J. Wanstall {edit.) A Darwin centenary p. 1 05- 1 20. London. MANTON, I. & REICHSTEIN T. 1961. Zur Cytologie van Polystichum braunii (Spenner) Fee und seinen Hybriden. Ber. Schw�tiz. Bot. Ges. 71: 370-383. METTENIUS, G. 1856- 1858. O ber einige Farngattungen (Fortsetzung). IV Phegopteris und Aspidium. Abhandl. Senkenberg. Naturforsch. Ges. Frankfurt a. M. 2: 285-420. MILDE, J. 1867. f.ilices Europae et Atlantidis, 127. Lipsiae. MI LDE, J. 1868; tl berAspidiu m nevadense Boiss. Bot. Zeitung 26: 360-364. MOORE, T. 1857-1862. 1ndex Filicum, London. MORTON, C.V. 1958. The Californian Species of Th elypteris. Amer. Fern J. 48: 136- 1 42. PANIGRAPHI, G. 1965. Preliminary studies in the cytotaxonomy of the Dryopteris villarii (Bell.) Woynar complex in Europe. Amer. Fern J. 55: 1-8. ROTHMALER, N. 1945. Der Formenkreis van Dryopteris paleacea (Sw.) Hand-Mazz. Ca ndo/lea 10: 91-101. ROUY, G. 1913. Flore de France, 14: Paris. ROY, S.K. & MANTON, I. 1965. A new base number in the genus Lygodium. New Ph ytol. 64: 286-292. 198 FERN GAZETTE: VOLUME 11 PART 2 & 3 (1975)

STAFLEU, F.A. et al. 1972. International Code for Botanical Nomenclature. Utrecht. STEARN, W.T. 1966. Botanical Latin. London & Edinburgh. UNDERWOOD, L. 1893. Our Native Ferns and their Allies, with synoptica/ description of the American Pteridophyta North of Mexico ad. IV. New York. VI DA, G. 1969. Tetraploid Dryopteris vil/arii (Bellardi) Woynar ex Schinz et Thell. in Rumania. /Bot KIJzlem 56 ( 1) 11-15. WI DEN, C.-J., VIDA, G., EUW J. VON & REICHSTEIN, T. 1971. Die Phloroglucide von Drropteris villarii (Bell.) Woynar und anderer Farne der Gattung Dryopteris sowie die , mbglichen Abstammu ng von D. fflix-mas ( L.) Schott. He/v. Chim. Acta 64: 2824-:-2850. WIDEN, C.J., FRASER-JENKINS, C.R., LOUNASMAA, M., EUW, J. VON & REICHSTEIN, T. 1973. Die Phloroglucide von Dryopteris caucasica (A.Br.l Fi"aser·Jenklns et Corley. He/v. Chim. Acra56: 83 1-838.

APPENDI X . Origin of plants used in the figures if not mentioned before :

TR- 746=Dryopteris filix·mas, Germany, Black Forest, Wiesental, between Zell and Schonau, c. 480 m alt., common, leg T. Reichstein, 7.1 0. 1.962. TR- 949=Dryopteris villarii subsp. pallida, Italy Mt. Bondone, N of Trento, bushy place above limestone c. 400-7QO m alt. rare, leg E. Hauser, 1962 cult. in Basel since 12.9. 1963. n=41 (G.V. 17.6. 1969) . TR-3558=Dryopteris abbreviata, Corsica, For�t d'A itone, along the road, c. 1330 m alt. common, leg H.L. & T. Reichstein, 25.6.1973, since then cult. in Basel. TR-356B=Dryopteris villarii tetraploid taxon, France, Caussols, N of Grasse1 deep cracks in limestone pavement, c. 1150 m alt. leg H.L. & T. Reichstein, 1.7. 1973 since then cult. in Basel. 2n=c. 160 (G. V. 26. 10.1973) . TR-3588=Dryopteris villarii subsp. pallida, Italy Abruzzi, Gale del Sagittario bushy place above limestone rock, c. 750 m alt. leg P. Brownsey, H. & K. Rasbach & T. Reichstein, 18. 7. 1973, since then cult. in Basel. TR-3623,;Dryopteris ablireviata, Corsica, Col de Taoria (N of Col de Verde) granitic block screec. 1600 & K. & T. 27.9. 1973, m alt. leg. H. Rasbach, H.L. Reichstein, · since then cult. in Basel. 2n=c. 80 (G.V., 27. 10.1973). TR-3628=Dryopteris abbreviata, Corsica, Pointe de la Monte, E of Col de Verde grantic block scree with Alnus viridis c. 1700 m alt. H. & K. Rasbach and H.L. & T. Reichstein, 28.9. 1973. CRFJ-21 26=Dryopteris villarii subsp. pallida, Turkey, Antalya, I rmasan Pass N of Akseki, c. 1500 m. alt. leg. C.R. Fraser-Jenkins, 21.8. 1910. n=41 (CRFJ 1.6.1974): No number in fig 6=Dryopteris filix ·mas, wild collection cu lt. in Base I.